[eug-lug]combing stdout stderr for tee
I am running a script on tcsh that calls 50 other programs. Execution takes about 20 minutes, so I want to see real-time what each program puts out to stdout and stderr when it exits. I also want to log stdout and stderr to a file. Some of the programs also require input from stdin. 'superScript | tee logFile' would be a good candidate to capture stdout, but it doesn't capture stderr. Q: is there a shell tool (for tcsh) that I could wrap around 'superScript' to combine stdout and stderr before tee'ing to logFile? Note, the clunky 2-terminal approach works for file rediretion A) 'superScript logFile' B) 'tail -f logFile' but I want to do better ... Horst P.S. I have tried aggregate commands using '()' and ';' but then the output is somewhat buffered and not contineous. ___ EuG-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [eug-lug]combing stdout stderr for tee
In reply to: Date: Mon, 23 Feb 2004 01:16:40 -0801 From: Jacob Meuser [EMAIL PROTECTED] ... 'superScript | tee logFile' ? I don't know tcsh, but the OpenBSD csh manpage says: The standard error output may be directed through a pipe with the stan- dard output. Simply use the form `|' instead of just `|'. Thanks for RTFM )-: it sure works !!! Actually I did a bit of rtfm, but didn't find anything searching for 'stderr', 'standard\ error' , 'combine' but to be fair the man pages say: Control structures should be parsed rather than being recog- nized as built-in commands. This would allow control com- mands to be placed anywhere, to be combined with `|', and to be used with `' and `;' metasyntax ((which I interpreted as the background...)) Diagnostic output may be directed through a pipe with the standard output. Simply use the form `|' rather than just `|'. - Horst ___ EuG-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [eug-lug]Not Linux Related, Assembly Related
Sorry folks, guess I'm a bit behind the times and kinda old school. I do No, you are not -- you just got hammered by a bunch of elitist techno-snobs who missed the actual point: menuetos (http://www.menuetos.org) is the ultimate anti-bloatware. A GUI, editor, games, assembler to add your own application, templates for half a dozen languages, plus more on just 1.4 MB (i tried it over a year ago so I forgot the additional stuff on it). Next time you may wanna say it fits on a 1.4 MB CD and you'll get much better press (I guess, I now have to study the procmail thread so I know how to handle all the subsequent hate mail :-) Off the subject, I still use floppies to write experimental LILOs to when I don't feel like tempering with a working machine, and to carry my phone/addressbook when I visit friends in Idaho with their '486 w/o modem, plus a few other things in between --but as said that's off the subject. - Horst ___ EuG-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [eug-lug]Not even *nix-related, asm is worthwhile
In reply to: Date: Wed, 18 Feb 2004 02:22:57 -0800 From: Ben Barrett [EMAIL PROTECTED] Reply-To: The Eugene Unix and GNU/Linux User Group's mail list [EMAIL PROTECTED] To: The Eugene Unix and GNU/Linux User Group's mail list [EMAIL PROTECTED] Subject: Re: [eug-lug]Not even *nix-related, asm is worthwhile Agreed. I overlooked menuetos until I finally got mention of the GUI and impressive features. The screenshots are amazing, really impressive stuff At least I didn't get shafted (-: going on there! My main concern is the hardware support: http://www.geocities.com/menuethw/byhardware.html ... I'm not clear as to whether the development on this has the floppy disk as a goal/spec/constraint, but I forsee a 50MB bootable mini-cd in the future that offers a knoppix-equivalent menuetos... and I mean equivalent by Na, I don't see a real application for such package on floppy -- however, putting it on Flash of a cheap SBC (single board computer) for a specific, designated application (tolerant against mechanical shocks) could be useful, you solve it for one set of hardware and there it does the job. But then, Flash is getting so cheap that you again can accomodate a larger OS... phew that helps a lot more: http://www.cs.helsinki.fi/u/twruottu/moshw/index.html YES, *another* guy (team) from Finland... If you actually look at the files in the package it's not just ugly binaries; it seems like some of the applications compiled from their .asm sources when needed. Neat and educational. Horst ___ EuG-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
[eug-lug]disable remote login for root?
How do I disable remote login (ssh) for root, and only allow 'su' or console login. It's a Red Hat box: [EMAIL PROTECTED] mail]# cat /proc/version Linux version 2.4.20-8 ([EMAIL PROTECTED]) (gcc version 3.2.2 20030222 (Red Hat Linux 3.2.2-5)) #1 Thu Mar 13 16:42:56 EST 2003 [EMAIL PROTECTED] mail]# Thanks . Horst ___ EuG-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [eug-lug]disable remote login for root?
It sure works (-: Ryan, Jason and Garl -- thanks for instant response. Ja, it's odd: the installation came with #PermitRootLogin yes which made me think the default is no. So there is no difference if you comment, or un-comment the Red Hat provided settings, hhm - whatever... I also saw other interesting stuff in /etc/ssh/sshd_config (-: - Horst In reply to: Date: Sun, 01 Feb 2004 22:20:37 -0800 From: Garl Grigsby [EMAIL PROTECTED] Reply-To: The Eugene Unix and GNU/Linux User Group's mail list [EMAIL PROTECTED] To: The Eugene Unix and GNU/Linux User Group's mail list [EMAIL PROTECTED] Subject: Re: [eug-lug]disable remote login for root? In /etc/ssh/sshd_config Change PermitRootLogin yes To PermitRootLogin no Garl horst wrote: How do I disable remote login (ssh) for root, and only allow 'su' or console login. It's a Red Hat box: [EMAIL PROTECTED] mail]# cat /proc/version Linux version 2.4.20-8 ([EMAIL PROTECTED]) (gcc version 3.2.2 20030222 (Red Hat Linux 3.2.2-5)) #1 Thu Mar 13 16:42:56 EST 2003 [EMAIL PROTECTED] mail]# Thanks . Horst ___ EuG-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug ___ EuG-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug ___ EuG-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [eug-lug]Hello, The Windows Virus
Hi Andrew -- good to see you here !!! All: Andrew is one of the youngest members on our list -- so be nice! About those 3 yrs ago I gave him a SUSE 5.2 pack I got from Jamie whom I had met at 5th St. Allen Brother 2(?) yrs before -- that's one of the places where EugLUG used to meet when it was still ELUG. So it's a nice local story, coming full circle :-) Andrew is the son of a friend, and at that time he was about 13 -- he walked away with the SUSE CD and figured out Linux all by himself, well almost (we spent a couple of hours on the phone, having fun). He is *the* next generation! --- Now, as for Win virus' I'd check first http://www.symantec.com/avcenter http://www.symantec.com/avcenter/vinfodb.html - Horst ~ In reply to: Date: Tue, 27 Jan 2004 19:57:23 -0500 (EST) From: Bill Essig [EMAIL PROTECTED] To: [EMAIL PROTECTED] Subject: [eug-lug]Hello, The Windows Virus Hello, first off, I am Andrew Wilson. Long ago, maby, err, 3 years ago, Horst got me into linux. I have been on a linux box ever (Mdk8 now) since, but I still own a MS windows box. I had recived word of a vius targeing domain owners, and I am one. I saw the email about it, but foolishly deleted it. I was wondering if anyone had any idea what it's all about. -- -Andrew Wilson a.k.a Bill Vinnegar Essig http://www.rpgopher.com ___ EuG-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug ___ EuG-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [eug-lug]new SCO DDOS virus
In reply to:
Date: Wed, 28 Jan 2004 14:09:32 -0800
From: Ben Barrett [EMAIL PROTECTED]
...
It is a .zip file, AFAIK.
Yup, I used 'strings' to inspect the baby I got, called document.zip in my
case, but names don't matter.
sdf:~:$ strings -n 2 document.zip |head -5
PK
0
document.scrMZ
PE
UPX0
# PK identifies the zip file
# .src -- expand to screen saver executables (I think?)
# MZ starts various DOS executables
---
sdf:~:$ strings -n 2 document.zip |tail -20
WH
wwx
wx
wx
;{w
KERNEL32.DLL
ADVAPI32.dll
MSVCRT.dll
USER32.dll
WS2_32.dll
LoadLibraryA
GetProcAddress
ExitProcess
RegCloseKey
memset
wsprintfA
PK
0
document.scrPK
*X
sdf:~:$
# the end of PK zip files have the names of files zipped up...
# ... here we see some potent Win32 .dll and more.
Cheers . Horst
On Wed, 28 Jan 2004 10:38:00 -0800
nyal [EMAIL PROTECTED] wrote:
| On Tuesday 27 January 2004 10:57 am, Nyal wrote:
|
| Just checked my mail and was also lucky enough to receive a copy. I'm
| using Kmail so when I saw the Hi subject I looked at the headers and
| everything in plain text.
|
| After the regular header type stuff is a whole boatload of fruit salad
| (letters, numbers, symbols, etc) no line breaks. Just curious as to what
| (if anything) all the mumbo-jumbo meant. Any ideas, O Great Gurus??
|
...
___
EuG-LUG mailing list
[EMAIL PROTECTED]
http://mailman.efn.org/cgi-bin/listinfo/eug-lug
[eug-lug]Off topic -- EFN related only
Sorry to bug this list, but this situation is quiet damaging to us(*), and I know there are folks on this list who can fix the problem in no time. (*)us being [EMAIL PROTECTED] (Cascade Canoe Club). We have sent out invitations to a couple hundred people for an upcoming event with references to our webpage hosted at EFN. But what they see instead is : You step in the stream, but the water has moved on. This page is not here. INSTEAD IT SHOULD MENTION THAT THERE IS A TECHNICAL SERVER PROBLEM, then people would come back, but now our organiztion looks pretty silly instead. Everybody understands that server problems and maintenance are unavoidable but the original warning (**) did not mention web outage: ~~~ Date: Wed, 21 Jan 2004 23:51:13 -0800 (PST) Planned Maintenance While EFN Moves Equipment On Sunday, January 25th, EFN system administrators will be moving some of our equipment out of our old machine room. Work will start about 06:00 and will continue until we have the moved equipment up and running at its new location. We hope to have things running by noon. During the outage EFN hosted services like email will not be available; dialup Internet service and DSL should continue to be available. Due to the changes in our service infrastructure this move will require, EFN dialup shell service will be unavailable for some time after the move is completed. While it is our intention to restore dialup shell when we can, we will not have the capability to deliver it, and we do not know when we will be able to do it again. We apologize for any inconvenience, and we thank you for supporting community networking! -- ~~~ - Horst ___ EuG-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
[eug-lug]SAO: TBP, Date Clarification, Thursday 1/15 (fwd)
Sorry if I am duplicating, but I didn't see the post in the archives. The TechBrew is always a good opportunity to network (employment and contract), maybe learn something new, meet new people, and to socialize with friends you otherwise only see on email .. Horst -- Forwarded message -- Date: Fri, 9 Jan 2004 08:19:35 -0800 (PST) From: Dave Clements [EMAIL PROTECTED] To: [EMAIL PROTECTED] Subject: SAO: TBP, Date Clarification, Thursday 1/15 Hello all, A slight correction to yesterday's e-mail. The Tech Brew Pub will be on Thursday, January 15 (not Thursday 1/16). I apologize for any confusion this may have caused. Please let me know if you have any questions. Thanks, Dave C Contents 1. A note from Matt about our NEW VENUE, and the upcoming year. 2. Tech Brew Pub on Thursday, January 15, at The Strand. 3. Job Announcements 4. Other SAO Lists New Year, New Venue === Yes, we are still here! As many of you may have heard by now, the Wild Duck closed its doors on January 1st. The SAO would like to thank the Wild Duck for their generosity and providing us with our meeting venue for the last 2 1/2 years. But life goes on. We have arranged for a new venue for our Tech Brew Pub meetings to be held at The Strand Building at 8th Charnelton. This is where Theo's Coffee House and Cozmic Pizza are located. When you attend you will notice a few changes, first of which there won't be any entrance fee. In the past the Wild Duck charged $5 which covered the food being served and staff, but, at least for now, admission will be FREE. Coffee, Pizza, Beer, Desserts etc. all are available, but if you don't wish to have any of that then attendance is FREE. Also, The Strand folks are providing our attendees with discount coupons for use during the meeting, so you can get some good food and beverages at an even better price! We plan to keep bringing you great technology speakers and content, informal networking, camaraderie and a fantastic opportunity to get out and meet with other IT professionals. Grab a friend and come on down and check out our new surroundings! Hope to Hope to see you there! Matt Jarvis SAO/ Lane Chapter President January SAO Tech Brew Pub = Topic: New Year, New Opportunities: A panel discussion on the current state of high tech employment in Lane County. When: THURSDAY, January 15, 5-7 pm Where: The Strand, 8th Charnelton, Eugene Cost: FREE Our presentation for January will be a panel discussion titled New Year, New Opportunities. We've assembled a group of speakers representing public and private organizations that work with technology professionals on workforce and career development issues. The panel will discuss the current state of employment opportunities in Lane County and will look forward through this year to identify where opportunities are likely to grow and how you should position yourself to take advantage of those opportunities. Paul Berger, from our SAO Chapter Leadership group, will moderate our panel that includes: Kristina Payne from the Lane Workforce Partnership, Diana Nicholas from Personnel Source, Larry Scott from Lane Community College, and Lee Jones from Lee Hechht Harrison. Mark your calendars and bring your questions. JOB ANNOUNCEMENTS = There are no announcements (yet!) this month. To get an announcement placed in next week's e-mail, please submit them by January 13. -- To get a job announcements posted here, please send the postings, as 20 lines or less of plain text, per job, to Dave Clements [EMAIL PROTECTED] -- Other SAO Mailing Lists === The state SAO office sends out a newsletter and a monthly calendar. If you are interested in receiving these please send an e-mail to Judy Cumby [EMAIL PROTECTED]. Corvallis also has an active SAO chapter that meets monthly. To get on the Corvallis list please go to http://www.peak.org/cgi-bin/majordomo?sao-corvallis:peak.org ___ sao_announce mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/sao_announce ___ EuG-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [eug-lug]any mandrake users? // Re: EuG-LUG Digest, Vol 10, Issue 11
I'm now reading my digest. Subject: Re: [eug-lug]any mandrake users? [EMAIL PROTECTED] Message-ID: [EMAIL PROTECTED] also if that fails and there is no libc.so.6 but there is a libc.so you can symlink libc.so.6 to libc.so ln -s libc.so.6 libc.so Shouldn't it be ? ln -s libc.so libc.so.6 This is just a note about the confusing 'ln' syntax; usually you execute cmd sourceFile targetFile but with ln that seem to be different. Larry knows... just one of the typos that slip in. - Horst ___ EuG-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re:[eug-lug]any mandrake users? (from the web archive)
I insyalled a long time ago so I don't remember details. But here is what I found.. Horst [root root]# cat /proc/version Linux version 2.4.19-16mdk ([EMAIL PROTECTED]) (gcc version 3.2 (Mandrake Linux 9.0 3.2-1mdk)) #1 Fri Sep 20 18:15:05 CEST 2002 [root root]# ls -ltr /lib/*libc* -rwxr-xr-x1 root root19436 Aug 19 2002 /lib/libcrypt-2.2.5.so* -rwxr-xr-x1 root root 1147848 Aug 19 2002 /lib/libc-2.2.5.so* -rwxr-xr-x1 root root 5088 Aug 29 2002 /lib/libcom_err.so.2.0* lrwxr-xr-x1 root root 13 Jan 16 2003 /lib/libc.so.6 - libc-2.2.5.so* lrwxr-xr-x1 root root 17 Jan 16 2003 /lib/libcrypt.so.1 - libcrypt-2.2.5.so* lrwxr-xr-x1 root root 17 Jan 16 2003 /lib/libcom_err.so.2 - libcom_err.so.2.0* [root root]# ___ EuG-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
[eug-lug]Comcast DNS related
I have been on ATT/Comcast for years. I was always in the 12.x.y.z IP block After being exclusively on Linux for quiet some time I just rebooted into Windows. After only 2 minutes of downtime (i.e. no lease-expire-situation) I got moved into the 24.x.y.z IP block, and my Windows Zonealarm spit out some warnings I am not expanding on here. Did anyone else on 'nix notice such a move into the 24.x.y.z. IP block recently, or am I just experiencing a random and normal change, or am I just a bit paranoid ? Just checking . Horst ___ EuG-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [eug-lug]dd / partition table / and restore on different device //
In response to: Date: Thu, 1 Jan 2004 16:22:22 -0800 (PST) Subject: EuG-LUG Digest, Vol 10, Issue 1 ... -- Message: 3 Date: Tue, 30 Dec 2003 23:26:20 -0800 (PST) From: [EMAIL PROTECTED] (Neil Parker) Subject: Re: [eug-lug]dd / partition table / and restore on different device // To: [EMAIL PROTECTED] Message-ID: [EMAIL PROTECTED] Content-Type: text/plain; charset=us-ascii Horst wrote, Q1: is the stuff after the MBR and below 7E00 hex the partition table ? The partition table is in that area, near the beginning. You should find the partition table starting at 0x1C0 and ending at 0x1FF from the beginning of the drive. There are four entries, each 16 (decimal) bytes long. ... Neil, thanks for expanding on this -- I am amazed to hear that the crucial info about how my hdb is divided into 13 partitions (ext2, ext3, FAT16, FAT32) --fits in just those 64 bytes ... (that must have been designed before bloatware, when 32kb of RAM were enough to go to the moon :-) ... Q2: if I'd just have a dd dump of some partition hdXN (like hdb1) how useful would this be if I try to restore it to another device? (assuming that one has enough space) I only recommend this if you restore to a partition that's the same size as the original parition. If you restore to a partition that's too small, the danger is obvious. If you restore to a partition that's too large, the restored partition may mount OK, but you'd only be able to access the restored size...the rest of the space would be unavailable. That would be OK if there are no other risks? -- as I recall, depending on partition tool and geometry, one doesn't always get the desired size so a small dead area at the end is no biggie if there is no other potential harm? - Horst ___ EuG-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
[eug-lug]dd / partition table / and restore on different device // EuG-LUG Digest, Vol 9, Issue 35
--- Bob Miller [EMAIL PROTECTED] wrote: Larry Price wrote: This is kind of a borderline question; A disk was intentionally zero'd out using dd if=/dev/zero of=/dev/hda however the DOS fdisk utility couldn't rebuild the partition table afterwards. This is shifting the subject a bit, but worth knowing: When I do: dd bs=100k count=1 if=/dev/hda of=hda_100k.dd dd bs=100k count=1 if=/dev/hda1 of=hda1_100k.dd (same for hdb and hdb1) and then use mc in hex mode to look at the four .dd files. I find that the hdX and hdX1 match at an offset of 7E00 hex (both drives are of different size and geometry) I see LILO in the MBR 0-1FF, then a few bits and lots of zeros in the range up to 7E00 hex. Q1: is the stuff after the MBR and below 7E00 hex the partition table ? Q2: if I'd just have a dd dump of some partition hdXN (like hdb1) how useful would this be if I try to restore it to another device? (assuming that one has enough space) Or in other words, is the stuff below 7E00 just 'metadata' concerning the organization of the drive as a whole with no relationship to the data that are on the individual partitions? In the past I have only used dd in the context of same_device - file transfer, but never across devices. Currently I have not enough room to test this in a safe way. Any feedback ? ... Horst ___ EuG-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [eug-lug]getting dumped
Quoting: From: Hal Pomeranz Date: Mon, 29 Dec 2003 18:25:29 -0800 ... Frankly, using GNU tar is probably a better option precisely because it doesn't have this particular issue (tar is a file at a time archiver, and is therefore relatively resistant to file system changes during the archive process). And GNU tar can deal with various weird files (sparse files, device files, etc.) that other versions of tar tend to have trouble with. I also like the above aspects of GNU tar. The only thing I found it doesn't like is tar'ing /proc/ (what's the point, anyhow?) and some stuff in /var/ that constantly changes. But for that I either use the -T option to select, or -X option to exclude. As for the single user mode: frankly, if there are frequent insert requests from httpd to /var/lib/mysql/someWebDB I wouldn't trust any binary 'snapshot' of an unflushed DB and use the utility that comes with the DB engine, like mysqldump; maybe I'm just paranoid, but the idea of 'better safe than sorry' is worth that little extra time; it's scripted anyhow. - Horst ___ EuG-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
[eug-lug]SAO: November TBP, 3 jobs at ORCAS, OPN raffle (fwd)
I haven't seen this posted yet -- sorry in case Im duplicating . Horst -- Forwarded message -- Date: Fri, 7 Nov 2003 13:39:30 -0800 (PST) From: Dave Clements [EMAIL PROTECTED] To: [EMAIL PROTECTED] Subject: SAO: November TBP, 3 jobs at ORCAS, OPN raffle Hello all, We have a number of things going on this month: 1. November SAO Tech Brew Pub on Wednesday November 12 2. OPN Computer Raffle 3. Three open positions at ORCAS Please let me know if you have any questions of comments, and I hope to see you on Wednesday. Thanks, Dave C --- November SAO Tech Brew Pub == Speaker: Michal Young, Head of the Department of Computer and Information Science at the University of Oregon When: Wednesday November 12, 5-7 pm Where: Wild Duck Music Venue, 6th Charnelton Cost: $5 at the door (cash only please) Michal Young, the Head of our own local computer science department will discuss what is happening in computer science at the University of Oregon. Michal will talk about several topics, including: - current research projects (and their tech transfer potential), - what students are learning these days, - how the department is involved with the community, and - what the future direction of the department is Michal has been at UO since 1997 and has been department head since 2002. Prior to returning to Eugene (he got his BS at UO), he was faculty at Purdue University. His area of research is software engineering with a particular emphasis on testing. Doors will open at 5:00, and the presentation will start between 5:30 and 6:00. OPN Computer Raffle === Oregon Public Networking (opn.org) has kicked off it's First Ever Computer Raffle! Help support community education through technology and get a chance to take home a sleek, jet black dream computer. Tickets are only $2 each, and it gives you a chance to win a top-of-the-line Pentium 4 (3-gigahertz) desktop PC with: 1 gigabyte of Double Data Rate Ram Radeon 9000 Pro video card 120 gigabyte hard drive 17 flat panel LCD monitor a DVD player and separate DVD burner Wireless keyboard and mouse Creative Labs 5-speaker set with sub woofer ATX Mid tower case with built in ethernet card This prize is valued at $3000.00 and can be viewed at the OPN office at 43 West Broadway, or on the OPN website- www.opn.org We are also offering a 2nd prize of free domain and web hosting with EFN for a year! Thanks to EFN for it's ongoing and generous support! Tickets are available from the OPN and efn staff members at our office in the heart of downtown Eugene. Call 484-9637 x113 for more information. JOB ANNOUNCEMENTS = Oregon Center for Applied Science -- a Eugene-based company that creates innovative life-enhancing multi-media programs for companies, employees and their families - has several job openings. Programmer 3: Programmer with expertise to write code used to develop, debug, and harden template-based programs aimed at users with widely diverse health needs. Our programs (delivered via the Internet, LANs and/or CD-ROM) typically use video presentations and data-driven content tailoring. Should be well-versed in PHP, Perl, CGI, HTML, Javascript, MS Access, SQL databases, mySQL, LINUX, Windows, and Mac environments. Must exhibit customer-service focus and ability to multi-task when working with team members who define program content and flow. Senior Programmer: Full or part-time senior programmer to provide outstanding technical and strategic leadership for in-house team of programmers. Candidate should have B2B expertise in designing range of programs including (a) highly-scalable, robust Internet-based content delivery systems for template-based, data-driven content (video-rich) including centralized authentication, administrative reporting, versioning; (b) CD-ROM programs including hybrid and/or micro server solutions; (c) sophisticated database architecture and network administration. Prior leadership roles at technology companies preferred (e.g., HP, Intel, Microsoft, Cisco). Must breathe technology trends and understand their relative value. Must be expert coder (proprietary open source) including Microsoft tools, PHP, mySQL, Perl, XML, CGI, XHTML, Javascript, MS Access, SQL databases, Windows, LINUX, and Mac OS. Candidate must lead by doing as well as by advocating, exhibit customer-service focus, and be able to multi-task when working with team members who define program content and flow. Macromedia Director Developer: Expert Director developer to work as contractor to help develop interactive instructional content. Participate with product development team to design presentation of defined content
[eug-lug] Red Hat Group in the Celebration Parade // the true reason...
Ben wrote: Date: Mon, 22 Sep 2003 10:19:19 -0700 From: Ben Barrett ... Subject: Re: [eug-lug]Red Hat Group in the Celebration Parade. Wow, I wish I would've known about that! Who did this? I finally had a chance to look at my Eugene Celebration photos, and found the true reason Ben wanted to be *in the middle* of that gorgeous Red Hat Society ! --see photo: http://wiki.euglug.org/index.php/SandBox BTW, did anybody see me in the parade? ...maybe not -- I walked in the shadow of the big ButtOn (again, see link above) BTW2, being in the parade is great fun; was my 4th time, and there should be penguins in it next year :-) - Horst ___ EuG-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [eug-lug]java
My overview is limited, but I know that a larger developer I worked for for over a year used Sun's JDK / JRE (basically as THE reference) From my own experience, I did some tests with Jython, cross compiling some python code to java classes, and then running those on different JVMs: the Kaffe that came with my Mandrake 9.x was not able to handle those. After I installed Sun's JDK / JRE it worked :-) ... Horst On Wed, 25 Jun 2003, Rob Hudson wrote: What's a good JDK to install on Linux? I think there are different varieties (Sun's JDK, IBM's Kaffe?, etc.) snip ___ EuG-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
wxPython // was Re: [eug-lug]python2exe
One note(downside) about wxPython: cuz it's based on C libraries it's
missing the nice, polymorphic function argument style of python ((in
contrast to TKinter :-) )):
func(defautlt1=val1, dflt2=val2,, **argDict{} )
where all you custom arguments are handled in one data structure
-argDict{}- optionally overwriting the defaults, and are passed down to
all the sub-functions.
Extremely convenient, not just for large GUI apps.. Horst
On Fri, 20 Jun 2003, Rob Hudson wrote:
Speaking of wxWindows. Here's an interesting project:
http://wxmozilla.sourceforge.net/
About:
wxMozilla is a project underway to develop a wxWindows component for
embedding the Mozilla browser into any wxWindows application. The
wxMozilla classes will utilize minimal, if any, platform dependant code
and utilize directly the XPCOM (Cross Platrom COM) interface to
instantate a html browser or editor within the application.
-Rob
On 20030619.1607, Ralph Zeller said ...
Python's great. It starts slower than many small unix programs but the
code is very readable. The wxPython library is a nice gui, there are
several ways to do gui development with wxPython--boa, pythoncard, wxGlade,
or vim!
... etc
___
EuG-LUG mailing list
[EMAIL PROTECTED]
http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [eug-lug]little programming challenge
On Fri, 20 Jun 2003, Neil Parker wrote: On Fri, 20 Jun 2003, Ben Barrett wrote: So, what is the problem we are really looking at? There are so many negative sentiments regarding Java, but I think it has its benefits, and is worth knowing what it is good for... right tool / right job, after all. snip The point, I think, is to illustrate how pointers are more restricted in Java than in C, and to provide fodder for those who like to complain about the inconvenience of languages that try too hard to protect you from yourself.[*] Sure, there is a price you pay for having (safer) references instead of plain pointers. (seat belts restrict your mobility) But what is the price we pay for having null-pointers, dangling pointers, buffer overflow ? --the hackers love it :-) snip - Neil Parker [*] See http://catb.org/esr/jargon/html/B/bondage-and-discipline-language.html Neil, are you sure this link addresses the above issue ? .. Horst ___ EuG-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
high- vs. low-level languages //Re: [eug-lug]little programmingchallenge
... I forgot to mention another feature of a high-level language that
helps with rapid development:
the garbage collector.
(which I first learned to appreciate in Java :-)
I still remember having to write full featured C++ classes with default
constructor, copy constructor (plus all the other overloaded
constructors), destructor, overloading assigment and other operators, and
on and on...
But why worry -- memory leaks only happen in theory, right?
- Horst
As variation for illustration:
~~~
#!/usr/bin/python
from sys import getrefcount
def swap(a,b):
return b,a, refcount in swap: + str(getrefcount('foo'))
a, b = 'foo', 'bar'
print refcount('foo') before swap, getrefcount('foo')
print swap(a,b)
print refcount('foo') after swap, getrefcount('foo')
~~~
On Sat, 21 Jun 2003, Horst wrote:
On Fri, 20 Jun 2003, Neil Parker wrote:
On Fri, 20 Jun 2003, Ben Barrett wrote:
So, what is the problem we are really looking at?
There are so many negative sentiments regarding Java,
but I think it has its benefits, and is worth knowing
what it is good for... right tool / right job, after all.
Sure, there is a price you pay for having (safer) references instead of
plain pointers. (seat belts restrict your mobility)
But what is the price we pay for having null-pointers, dangling pointers,
buffer overflow ? --the hackers love it :-)
snip
___
EuG-LUG mailing list
[EMAIL PROTECTED]
http://mailman.efn.org/cgi-bin/listinfo/eug-lug
[eug-lug]sign up soon / Python for Fun Syllabus
All -- those of you interested in Chris Meyers' class should sign up very sn. Should you have any questions feel free to contact Chris by email or by phone (see below) - Horst Course Information Python for Fun Summer Term, 2003 Thursdays 6pm to 9pm; Jun 26 to Aug 14 DTC room 124. Instructor: Chris Meyers. 343-4349 [EMAIL PROTECTED] Course Description. An introduction to programming and to the Python language and its vast library of extendable modules. Covers the basics of programming as well as a offering a look at the power of todays modern languages. Course Objectives. By the end of the course you should be - acquainted with how computers are basically programmed - familiar with the core of the Python language and how it compares with other languages such as C, Basic, Java, and Javascript - able to write and debug small programs - comfortable exploring Python on own using online sources and literature. Class Format: Each class will meet for 3 hours with a 15 minute break about midway. The course will be a combination of lecture and hands on with Python on the computer along with some challenge problems for homework. Required Materials: None, but I will recommend some books that are available. Much of the course will be structured around the online book How to Think like a Computer Scientist which may be freely printed out and copied. Disability Services: If you are in need of academic support because of a documented disability (whether it be psychiatric, learning, mobility, health-related, or sensory), you may be eligible for academic accommodations through Disability Services. Contact Disability Services at 747-4501, ext. 2150 or TTY 741- 3079, or stop by our office in Center 217. Week 1: The nature of computation. What do computers REALLY do? The nature of algorithms. nesting, looping, recursion The 1st program. Compared with Java, C, C++ Simple data values. int, long, float. operators (+,-,*,/,**,%) Boolean and string values. An introduction. variables and namespaces Week 2: Functions. Built in. (type,len,id,int,float,str). Coercion User written functions. indentation. Flow of execution. Stack Diagrams Parameters and arguments. Return values, Local vars Conditional execution. if, elif, else 1st real-world function. tax.py First recursive functions. Avoiding infinite recursion Iteration with while Week 3: Compound data items. Strings again. sequences and slices. String formatting operations with % For loops iterate over sequences. range function string module. 1st import. Example functions with for, while, and recursion. Lists. sequences, slices, and membership. + and * ops Week 4: Mutable objects. As parameters to functions. cloning List methods. len, sort, reverse, etc. Tuples are immutable lists. Tuple and list assignment Dictionaries. initializing, modifying. Methods for dictionaries: len, keys, values, items, has_key, get, copy Uses for dictionaries. lookups, existence, secret lives. Week 5: File I/O. How files are organized. File operations: open, close, read*, write*, seek. Example programs to copy and filter files. Python exceptions. Traceback. try, except, raise. Using pickle module to save Python data to disk. A simple report program Week 6: Objects. Simplest class definition. Add attributes by setting them. Objects and methods. __init__ constructor method Objects and inheritance. Play with animal classes. Week 7: Exploring the Python library. string, re, glob, httplib, getopt,random,socket,webbrowser Week 8: basic graphics. drawing shapes bouncing balls, tower of hanoi Gui programming. Tk widgets -- ___ EuG-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [eug-lug]sign up soon / Python for Fun Syllabus
On Thu, 19 Jun 2003, Rob Hudson wrote: Do you know how much it is? As I recall from the winter term it was close to $100 (though I may be off by 10 bucks). I don't have my LCC booklet handy, but the LCC website offeres signup with credit card, or a 24/7 automated phone line to sign up and/or to get the amount. Horst On 20030619.1745, Horst said ... All -- those of you interested in Chris Meyers' class should sign up very sn. Should you have any questions feel free to contact Chris by email or by phone (see below) - Horst Course Information Python for Fun Summer Term, 2003 Thursdays 6pm to 9pm; Jun 26 to Aug 14 DTC room 124. Instructor: Chris Meyers. 343-4349 [EMAIL PROTECTED] ...etc. ___ EuG-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [eug-lug]little programming challenge
On Thu, 19 Jun 2003, Larry Price wrote: Neil popped this one on the clinic, can you write a function in java that does the equivalent of this python script ---snip #!/usr/bin/python def swap(a,b): return b,a a = 'foo' b = 'bar' print a,b a,b = swap(a,b) print a,b --snap- turns out to be surprisingly difficult ;-) Cool :-) The convenience python offers here doesn't really come from the swap function, but from the transparent 'tupling' in-and-out -- at both sides of the assignment operator, and at the return. Horst (No, I won't show the vector in-and-out version in Java, neither in C++ , for that matter :-) #!/usr/bin/python def swap(a,b): return b,a a, b = 'foo', 'bar' print a,b a,b = swap(a,b) print a,b print return type of swap():, type(swap(a, b)) ___ EuG-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [eug-lug]python2exe //Re: EuG-LUG Digest, Vol 3, Issue 46
Date: Thu, 19 Jun 2003 14:41:37 -0700 From: Rob Hudson [EMAIL PROTECTED] Message-ID: [EMAIL PROTECTED] How is the executable on speed? Does it start up pretty quickly? Yes, pretty fast. I played with small console apps about half a year ago. The shared python22.dll is about 800kb, and the actual .exe files around 50kb. - Horst ___ EuG-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
[Eug-lug]external watchdog ?
Does anybody have experience with, or suggestions about external watchdogs? We are having a pesky hardware problem that occasionally freezes a Sun box, leaving us with the power switch as the one and only control 'device'. Until that's fixed someone has to get there physically, and flip the switch. From my embedded systems days I remember on-board watchdogs (which we don't have here), so my idea was find some sort of external device. I am not sure which kind of I/O and control would be best for a 'nix box but I can envision a daemon that is activated/deactivated during init/shutdown(or forced reboot), controlled via serial port, NIC, __?__ , ... Maybe I'll just leave it at that point -- any ideas/experiences ? - Horst ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]tonight clinic / another(?) java ?
Thanks to all for sharing related Jython/Java experiences, feedback on re-inventing the wheel thoughts and such... To clarify, and to close the subject. Running Jython on Sun's JVM was no problem, but what I was interested in was exploring to which extend I can port (jythonc) *some* python program so that it can be run on *any* client's JRE. (*some* and *any* left intentionally vague) Unfortunately, that didn't go as far as I was hoping :-( - Horst On Fri, 28 Mar 2003, john fleming wrote: Horst wrote: I am sure I am missing something, and 'kaffe' could be trained, but why? Why re-inventing the wheel ? - Horst When I was installing jython it sais in the docs that sun's JRE was the best bet and being both stubborn inquisitive I tried kaffe amd blackdown and sun was the only one that worked.So for this specific purpose I wouldn't mess around anyone have any experiences with other java implementations? ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]tonight clinic / another(?) java ?
On Thu, 27 Mar 2003, Larry Price wrote: ... also java on debian write once debug everywhere? what is this ? -- why don't distros just add a link so users can download Sun's SDK and JRE ? -instead of re-inventing the wheel ? Is this a licensing issue ? A couple of weeks ago I explored Jython as a way to port python programs to run on the JVM; worked OK(ignoring a couple of flaws) until I tried it on my Mdk 9.0 distro which comes with 'kaffe' instead of the real thing. I am sure I am missing something, and 'kaffe' could be trained, but why? Why re-inventing the wheel ? - Horst ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]binhex --via python :-)
On Tue, 11 Mar 2003, Rob Hudson wrote:
Whenever someone sends me a file from the mac, it appears garbled. An
example header looks like this:
(This file must be converted with BinHex 4.0)
:$%KPE(!JBQN,R4iG!849K8C'pcB3!X)3$M+ca)9%e-2Jd+2%K383
q$3Sm9%P86%8q5'9XF$`[9%P86%8q$3Sm6%P15b548`p)[EMAIL PROTECTED])'K
[EMAIL PROTECTED])Q0cFbpcG(PXCA-ZBh0c)LdHAP25*dCAKd,f0cFb)q$3SmE'PZDb[EMAIL
PROTECTED]
...
Is there a way to un-binhex this on Linux?
PYTHON is your friend; platfom independent, basically just two lines of
code:
~
from binhex import hexbin
hexbin('inName', 'outName')
~
Or you could also run it on the fly via command line... Horst
python -c 'from binhex import hexbin; hexbin(inName,outName)'
FROM THE DOCS:
#
12.14 binhex -- Encode and decode binhex4 files
This module encodes and decodes files in binhex4 format, a format allowing
representation of Macintosh files in ASCII. On the Macintosh, both forks
of a file and the finder information are encoded (or decoded), on other
platforms only the data fork is handled.
The binhex module defines the following functions:
binhex(input, output)
Convert a binary file with filename input to binhex file output. The
output parameter can either be a filename or a file-like object (any
object supporting a write() and close() method).
hexbin(input[, output])
Decode a binhex file input. input may be a filename or a file-like object
supporting read() and close() methods. The resulting file is written to a
file named output, unless the argument is omitted in which case the output
filename is read from the binhex file.
#
___
Eug-LUG mailing list
[EMAIL PROTECTED]
http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]Mini Knoppix
Dave, I downloaded the ISO (was extremely slow) and burnt a CD for you which I left with Larry at the EFN. The iso is a bit over 200 MB. In case some one else is interested let me know and I'll make another copy. BTW, the author, Klaus Knopper, is German, so the K in Knoppix is *not* silent (just as a tip to the presenters --you have about a week to practice :-) ... Horst On Thu, 6 Mar 2003, Dave Wyatt wrote: I found this a while back. I never had the chance to download it but maybe it will be usefull to some of you. ftp://ftp.es.debian.org/pub/miniKnoppix/knoppix_lite_1.0_EN.README The ISO is ftp://ftp.es.debian.org/pub/miniKnoppix/knoppix_lite_1.0_EN.iso __ Do you Yahoo!? Yahoo! Tax Center - forms, calculators, tips, more http://taxes.yahoo.com/ ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]user webpages on my box
directory you have to add this. That's correct. mkdir ~/public_html; chmod 755 ~/public_html And if you don't want to have visitors sniffing your entire ~/public_html/ directory/file structure but rather limit them to the html files you decide to present then chmod 711 ~/public_html However, then you DO need some index.xxx xxx :: htm html shtml cgi or whatever is defined in httpd.conf (I think the presidence is alphabetical or defined in the conf file (?) - Horst ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]user webpages on my box // perms for directories
On Sat, 15 Feb 2003, Timothy Bolz wrote: I found the solution after search the web. The one site said I have to have chmod o+x ~ of my home directory. It worked and now the webpages come up. Maybe someone can enlighten me on why this worked. Wouldn't this be a security breach if anyone can execute code? No, permissions for directories work differently. o+x (or g+x) means that the world ( or group) can 'use' files in that directory. The meaning of 'use' is determined by the permissions of the actual file, the usual three r, w, or x o+r means that the world can read(list) the directory (if they can read, edit, or excecute any file in that dir depends again of the file's perms) o+w means that the world can change a file's name (if that file allows it) To memorize this it helps to look at the directory as just another 'file', sort of the listing of what's in it. From that perspective: x = use what's in it r = list what's in it w = change the listing, i.e. the name of a file in it (again directory perms are just a pre-filter -- the perms of the actual files in it decide).. Horst ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]James Gleick on Spam / (Limit to 4 email?)
On Tue, 11 Feb 2003, Bob Crandell wrote: ... alot. My thought is that SPAMers are using technology to create a problem therefore we should use technology to frustrate their efforts. No technology and no government oppression is going to stop SPAM just like it didn't stop junk FAX. But Bob, thinking as a business owner - if spammers intentionally waste your time and money (as it has happened in the past, and maybe even related to your current thread regarding Limit to 4 email (*), why then shouldn't you be legally entitled to recover some of the costs, as for any other intentional damage inflicted on your business? For that you need some legal leverage. Those uninvited IP packets on your networks and machines are a form of modern trespassing. Sure, it won't *completely* solve the problem, like laws against trespassing haven't *completely* eliminated trespassing --but do you wish the government would have left it all just up to you to deal with trespassers? ...Horst (*) even if SpamAssassin isn't the problem causing Limit to 4 email you are spending the time investigating it -- w/o the proliferation of spam you wouldn't have to :-) ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
nobody's shell // was Re: [Eug-lug]mail to root under postfix ?
1) Talking about nobody -- what should be his/her shell ? My distro decided on /bin/sh whereas some system accounts are assigned to /bin/false (or true), neither giving them much power. I guess I could try and wait until some application using nobody breaks... but maybe you folks wanna spoil that sort of learning experience??? 2) What is the recommended entry for the 2nd field (password indicator) for users that can't log in (like my proxie for root mail) ? My guess is 'x' in passwd and '!!' in shaddow -- whereas Mandrakes utilities decided on '*' in passwd and didn't make and entry in shadow(!?) 3) HowTo? Normally I use the general tools useradd, usermod and not some distro-specific tools -- any comments? (other than manually editing the files in /etc/ -which I used to do in the past) grep nobody /etc/passwd nobody:x:65534:65534:Nobody:/:/bin/sh grep nobody /etc/shadow nobody:*:12069:0:9:7::: grep postfix /etc/passwd postfix:x:77:77:system user for postfix:/var/spool/postfix:/bin/false grep postfix /etc/shadow postfix:!!:12069:0:9:7::: - Horst On Tue, 4 Feb 2003, Jacob Meuser wrote: On Tue, Feb 04, 2003 at 07:07:24AM -0800, P Casper wrote: It's not much less clunky, but as a newbie, I let root's mail go to the default 'user' == nobody and created an alias in .bashrc so that I could pull up mail for that user without much trouble. *shrug* It works. Be careful about 'nobody'. It can mean 'everybody'. For example, services than anyone can use sometimes run as 'nobody' (ie, the service owns root's mail), and you probably don't want to make it any easier for 'everybody' to read root's mail, as it may contain sensitive info. ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]mail to root under postfix ?
On Tue, 4 Feb 2003, Darren Shepard wrote: ... Any mail client (gui or not) can have security bugs (eg. pine has a history of exploitable buffer overflows). ... Convinced ! -- Thanks for clarifying :-) Also thanks to Cory for the link. Indeed, http://www.postfix.com/faq.html is very well organized and maintained. - Horst ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]Re: Freaky and Naked on the Net
Ben, you put together some interesting thoughts and links (I just finished the motherjones article and had to resist the temptation of quoting about half of it right here. Instead, let's get concrete and ask the best ISP in town (EFN) a few questions: ~ for EFN ~~~ - Have your records, logs, etc. ever been subpoenaed ? - Are you legally able, and willing to comment on such issues ? (I understand, that's pretty generally phrased (I am not a lawyer) -maybe EFN has some general policy on the subject that is public ? ) - For how long are dialup logs kept (the userID-dynIP-timestamp connection) ? end for EFN I don't question that there are reasons to subpoena some of the information we are talking about here under *specific* circumstances (e.g. the fairly recent bust of a cyber child-porn ring in our area) -in contrast, the Poindexter project is just a big fishing expedition. One comment on the motherjones article regarding google (towards the end of the article). Barlow may be right in judging the past and current situation. But then, so far we have only seen the episode 'Google -- part I'. Let's not forget, they are running an efficient business, and when they spend the money on collecting and storing all the client requests ever made together with all the client info they can get, one can only speculate what the business model for 'Google -- part II' may be like... - Horst On 5 Feb 2003, Ben_ wrote: I'm glad you posted the google-watch links, that's very interesting information. I am concerned by the popular belief that IP addresses somehow identify a single computer, however!!! In many cases, a single IP address is shared (say, via NAT) by dozens or even hundreds of computers. It can also be spoofed. I do understand that in certain instances it can be used to establish legal (ie, criminal) identifiability, though I do find this troubling. This bit, from the krane.html, bothered me in particular: The fact that you record unique cookie ID, plus IP number, plus date and time, makes much of your information identifiable. Authorities can also do a sneak and peek search of a Google user's hard drive when he isn't home, retrieve a Google cookie ID, and then get a keyword search history from you for this ID. Are they really suggesting that authorities will physically enter, JUST to get a google cookie ID off a user's hard drive, and then return to google to find out what that user has been searching for?? If they go to the trouble of physically searching the hard drive, why would the cookie ID be more useful than all the other data there?! Or are they suggesting that they remotely access a user's hard drive? Yeah, right -- that's the M$ NSAKEY backdoor, right. Just like John Perry Barlow reiterated in his recent motherjones interview [1], the TIA [2] project mostly offers LOADS more haystacks, and not many needles. I would personally suspect that anyone with something to hide, something valuable anyway, would take efforts to cover their backs. Look at how hard it has been to unravel the Enron and other accounting scandals! There are plenty of web-anonymizing services, and clever uses of proxies, etc, to allow for instance dedicated information specialists in communist regimes to freely access otherwise-blocked information (thanks in part to CDC [3] and other hactivists), that methinks only the dumbest criminals would be convicted based on google's secret data. So yes, I agree that it is futile for google to abuse the general user's privacy to this extent. As far as google seeking a deployment engineer with gov't security clearance, that makes sense to me!! I know a lot of organizations deploy google on their LANs [4], and I'd assume that some of them indeed need someone with clearance to simply be able to walk in their doors, and do anything at all on their network (based on their IT/security policies)... this does not simply imply federal/military ops, but also the whole slew of privately-owned contracting businesses that work indirectly for said ops... again, this is not limited to military, either -- keep guessing. I don't see anything inherently evil in those policies, although it certainly makes it difficult to public AUDIT those organizations and operations! For those of us here, who fall into the category of freedom zealot, please remember that secrets are still important in a free world!! The freedom of secrecy must be respected, even if you don't agree with others' motivations. ciao. Ben B On Tue, 2003-02-04 at 17:32, Horst wrote: And to add another aspect... Maybe your privacy is less threatened by individuals such as the one attempting step (1)-(4) than BigBrother himself: See quote below from http://www.google-watch.org/krane.html They ask some reasonable questions: - Why does google need a cookie that doesn't expires before
Re: [Eug-lug]mail to root under postfix ?
On Mon, 3 Feb 2003, Darren Shepard wrote: On Mon, Feb 03, 2003 at 03:24:46PM -0800, Horst [EMAIL PROTECTED] writes: ... | | ~~~ from the /etc/postfix/aliases ~ | ... | # For various security reasons, postfix WILL NOT deliver mail as root, so | # ensure that the root alias is aliased to a HUMAN user, as otherwise | # mail may get delivered to the $default_privs user (nobody). | ... [snip] ... It's a bad idea for root to recieve mail directly (ie. because you then invoke a mail client as root, and read a spool file full of potentially malicious data from the net). Interesting. Could you expand a bit on that ? I am using a text based mailer, and almost never GUI as root. As I said earlier, I like root to handle internal administrative tasks, and recieve reports generated by root's own cron jobs (I'd consider that a clean concept since that's what root is for, though security issues add another twist to it). So a) creating an admin user that receives root's mail b) an alias for root, like pine='su - admin' c) last step of login script for admin opens pine d) upon pine exit admin logs out would make it both, smooth and safe, I guess. - Horst ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]Naked on the Net (you guys already KNEW this!)
And to add another aspect... Maybe your privacy is less threatened by individuals such as the one attempting step (1)-(4) than BigBrother himself: See quote below from http://www.google-watch.org/krane.html They ask some reasonable questions: - Why does google need a cookie that doesn't expires before 2038? - Why logging the originating IP (i.e. if you just throw your cookies away while having a (semi)static IP, an 'unusual' OS plus browser, still makes you a pretty unique client) And on http://www.google-watch.org/jobad.html - Why does google want to hire a Deployment Engineer (with security clearance): Must have current government top security clearance (TS/SI). Google currently does not allow outsiders to gain access to raw data because of privacy concerns. Searches are logged by time of day, originating I.P. address (information that can be used to link searches to a specific computer), and the sites on which the user clicked. People tell things to search engines that they would never talk about publicly -- Viagra, pregnancy scares, fraud, face lifts. What is interesting in the aggregate can seem an invasion of privacy if narrowed to an individual. So, does Google ever get subpoenas for its information? 'Google does not comment on the details of legal matters involving Google,' Mr. Brin responded. -- New York Times, 28 November 2002 On Tue, 4 Feb 2003, Harald Sundt wrote: Can it REALLY be this easy? Hal Quote: I can't see the posts you're responding to (since I surf here through Google Groups), but if some troll doesn't understand the need for privacy, it's easy to convince them otherwise. Here's how to do it: 1) Get the person's IP address from their post's headers, then trace the IP addy to the Internet Service Provider. 2) Posing as an attorney or police detective, call the person's Internet Service Provider and ask them who used that IP address. 3) At this point, the idiot employee of the ISP usually tells you everything he knows. - But if he doesn't tell you, then all you have to do is mail certified a fake subpoena to the ISP and then get the records that way. (Almost nobody knows what a real subpoena looks like, so it's easy to fake one.)- 4) Once you have the person's name and address, you can do a public records search for their Date of Birth. Also, the Social Security Number usually appears in one record or another. (Depending on the state they live in, their driver's registration records might reveal it for example.) 5) Then post on Usenet the person's name, address, date of birth, Social Security Number, and driver's license number. 6) The final result is that the person will then realize how important privacy is. :) :Unquote ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
[Eug-lug]mail to root under postfix ?
I am used to sendmail, but mdk 9.0 defaulted to postfix, so I decided to try it; i.e. I am asking a postfix-newbie question. I noticed in postfix's alias file that mail to root will be only delivered to a regular user (see clips below) The most transparent work-around I found 1) creates a proxy user, e.g. 'rootx' . 2) set that in 'aliases' 3) point mail program's INBOX to /var/spool/mail/rootx Am I missing something? --how do others handle this less clunky ? - Horst ~~~ from the /etc/postfix/aliases ~ ... # For various security reasons, postfix WILL NOT deliver mail as root, so # ensure that the root alias is aliased to a HUMAN user, as otherwise # mail may get delivered to the $default_privs user (nobody). ... # Person who should get root's mail. This alias # must exist. ## default::root: postfix ## goes to nobody:: root: root root: rootx ~~ ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]mail to root under postfix ?
On Mon, 3 Feb 2003, Bob Miller wrote: Horst wrote: I noticed in postfix's alias file that mail to root will be only delivered to a regular user (see clips below) ... Am I missing something? --how do others handle this less clunky ? Don't you want root's mail delivered to your own mailbox? Wouldn't it be easier to read all your mail from the same box? Actuall, for internal administrative tasks, root's cron jobs, etc. I like to keep the mail in root's mailbox. It seems like exim handles it similar (except that postfix prefers its own /etc/postfix/aliases --and refuses to deliver to root directly) Thanks Bob . Horst ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug] archive option(?) / Free Wireless
On 2 Feb 2003, Ben Barrett wrote: snip I heard XYZ Street (the pub/restaurant) has an open access point in the vicinity -- hope it stays up now that this message hit the list archives!! ... I was always wondering how difficult it would be to implement on option [keyword in subject line] that turns a message into a send-to-list but dont-archive option; this keyword would have to make it into one of the optional fields of the mail header for subsequent thread handling. It's not just about 'secret' access points, but as others have pointed out, the archives are one of the top level windows into EugLUG. There is lots of good stuff on the list (I make my own archive once a month) but on occasion there are some pretty goofy and better-not-archived threads - many of which are perfectly fine for local communication, but do you always want google to add this to your eternal online resume ?... - Horst ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
[Eug-lug]UO Colloquium, Thursday, January 30 (fwd)
Maybe interesting to some us. Sorry for the short notice ... Horst -- Forwarded message -- Date: Wed, 29 Jan 2003 10:37:58 -0800 From: Cheri Smith [EMAIL PROTECTED] To: [EMAIL PROTECTED], [EMAIL PROTECTED], [EMAIL PROTECTED], [EMAIL PROTECTED] Subject: UO Colloquium, Thursday, January 30 Why IPsec is Problematic for Applications other than VPN's Jonathan Trostle Security Consultant ABSTRACT This talk focuses on the Internet Protocol Security (IPsec) policy mapping problem: given an outbound packet originated on the local host, how does the network layer decide if IPsec should be applied to the packet. More generally, how does one ensure correct mappings between application level identifiers and network layer identifiers. Although DNSSEC is one tool that can be applied here, it is neither necessary or sufficient. We describe some results for new techniques that that can be used for some legacy applications to partially or completely solve the IPsec policy mapping problem. We show how the set of current IPsec policy parameters can be usefully expanded. Finally, we summarize the environments today where IPsec is being used and discuss which IPsec policy mapping techniques are most appropriate for these environments. DATE: Thursday, January 30, 2003 TIME: 3:30 p.m. talk, refreshments following talk PLACE: 220 Deschutes Hall (Colloquium Room), University of Oregon *** Cheri Smith Undergraduate Coordinator Computer Information Science Phone: (541) 346-1376 1202 University of Oregon E-Mail: [EMAIL PROTECTED] Eugene, OR 97403-1202 Fax: (541) 346-5373 Office Hours: Monday through Friday, 9:00 a.m. - 12:00 p.m., 1:00 p.m. - 5:00 p.m. *** ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Mdk 9.0 // was Re: [Eug-lug]Mandrake 8.2 hangup ....
Dave, that would be great! If you could drop them off at EFN I'd return to you - yes in a timely manner (address and more details off the list) I will also make copy for Jamie. If there is a meeting tonight I'll pick them up there, if not by Thu. night Horst On Wed, 15 Jan 2003, Dave Wyatt wrote: I have the 3 CDs for Mandrake 9.0 I can drop off for you. Since I was laid-off I have had to stop my Cable connection so I can't get any more ISO images (at least not in a timely manner). Man, dial-up sucks after you've had a nice fast connection. I seem to have lots of time lately so I can get these CDs to someone today. I just ask that they be returned after copied or used. Dave --- Linux Rocks ! [EMAIL PROTECTED] wrote: Sadly, I dont have a copy. Tim made me a copy, but I ...etc ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
[Eug-lug]Mandrake 8.2 hangup (?in? X11 / KDE 2.2.2)
I am reposting this with additional info on X11. I had this problem since I changed from mdk 8.1 to mdk 8.2 (2.4.18-6mdk). 8.1 ran w/o any problems for more than half a year on the same hardware. Problem: I start X11/KDE 2.2.2 as normal user, opening a few consoles in KDE, maybe a browser; suddenly certain processes stop working - most often observed in consoles, 'cuz most used. Even switching out of X to text console doesn't stop this. I start looking at various status information; soon all is frozen, leaving me with very few chances to do any real-time diagnostics. Shutdown hangs too, very early on, or doesn't even start when I try too much before shutdown. 'ps -ef' hangs between X11 and KDE start (!) -see clips below. CPU load goes up to 5.x for no (good) reason. I had that after I installed mdk 8.2 on my original, previous ext2 partition (that installation got trashed at one of those crashes). And now, with a new installation on ext3 I see similar behavior again. I didn't see any error message in the logs -but it's possible that I overlooked some subtle information that points at a problem ? Any pointer as to what to look for in which logs? Any tips or comments? .. Horst (I looked at the Mandrake archives where someone posted an *unanswered* message regarding memory leak under X, almost a year ago... doesn't fit exactly my situation (as far as I understand the post)... hard to imagine I am the only one since that distro came out...) == ps -ef = UIDPID PPID C STIME TTY TIME CMD ... root 1924 1 0 00:03 vc/5 00:00:00 login -- horxx root 1925 1 0 00:03 vc/6 00:00:00 /sbin/mingetty tty6 root 1928 1920 0 00:03 vc/1 00:00:00 -bash horxx 2018 1924 0 00:03 vc/5 00:00:00 -bash horxx 2044 2018 0 00:03 vc/5 00:00:00 /bin/sh /usr/X11R6/bin/startx horxx 2055 2044 0 00:03 vc/5 00:00:00 xinit /etc/X11/xinit/xinitrc -- -deferglyphs 16 root 2056 2055 4 00:03 ?00:01:04 /etc/X11/X :0 -deferglyphs 16 ==CONSOLE HANGS HERE FOR EVER== == next line would be normally: (different PIDs at diff. boot)== horxx 2235 2229 0 01:06 tty5 00:00:00 /bin/sh /usr/bin/startkde ... == top: = 12:29am up 26 min, 3 users, load average: 4.88, 3.37, 1.90 93 processes: 91 sleeping, 2 running, 0 zombie, 0 stopped CPU states: 2.7% user, 1.5% system, 0.0% nice, 95.6% idle Mem: 255772K av, 246160K used,9612K free, 0K shrd, 46792K buff Swap: 546168K av,3032K used, 543136K free 55928K cached PID USER PRI NI SIZE RSS SHARE STAT %CPU %MEM TIME COMMAND 2056 root 15 0 26344 21M 2548 S 2.1 8.5 1:08 X 2288 horxx 11 0 9304 8680 7316 R 0.9 3.3 0:05 kdeinit 10498 horxx 15 0 1064 1064 812 R 0.7 0.4 0:00 top 2272 horxx 9 0 9268 9268 8416 S 0.1 3.6 0:01 kdeinit 2285 horxx 9 0 9148 9148 8392 S 0.1 3.5 0:00 kdeinit 1 root 8 0 504 504 440 S 0.0 0.1 0:04 init 2 root 9 0 00 0 SW0.0 0.0 0:00 keventd 3 root 9 0 00 0 SW0.0 0.0 0:00 kapmd 4 root 19 19 00 0 SWN 0.0 0.0 0:00 ksoftirqd_CPU0 5 root 9 0 00 0 SW0.0 0.0 0:00 kswapd ... == the resources above used by X are somewhat more than I see under redhat 8.0 == but that doesn't correlate with a CPU load of almost 5 ??? == X11 = XFree86 Version 4.2.0 / X Window System (protocol Version 11, revision 0, vendor release 6600) Release Date: 23 January 2002 If the server is older than 6-12 months, or if your card is newer than the above date, look for a newer version before reporting problems. (See http://www.XFree86.Org/) Build Operating System: Linux 2.4.18-1mdksmp i686 [ELF] Module Loader present ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]Mandrake 8.2 hangup (?in? X11 / KDE 2.2.2)
On Tue, 14 Jan 2003, Linux Rocks ! wrote: Horst, Id suggest trying different ram in the computer for a few days and see if the errors continue. often ram can seem fine, but actualy be bad, and cause weird errors like you are seeing. Hhm, to me there seems to be too much of a specific pattern (X11, KDE, with just that distro) than I'd blame it on the RAM (I also don't have the spare )-: Looks more like something is gradually 'filling' up, no 'bang', core dump, etc... i.e. the memory leak post sounded more appealing. Also working on this problem is no fun since a corrupted file system is always a possibility (happened once) - i.e. I'v been just avoiding X. Also, mandrake 9 has been out for a while now, you might try it and see if the problems still continue (since mandrake Now, that's a thought ! Where/How to get ? -I am happy to trade blanks or burnreturn :-) ^^^ - Horst 8.1 worked, and the problem has only happened in 8.2, its possible that the issue is specific to 8.2. Jamie On Tuesday 14 January 2003 10:51 am, Horst wrote: : I am reposting this with additional info on X11. etc... ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]uh-oh disk drive crash?
On Sun, 12 Jan 2003, Rob Hudson wrote: ... I'm not exactly looking forward to the extraction process, but I'm going to try to load up on knowledge before hand and see what I can salvage (if anything). Rob, it sounds like bits and bytes *continue* crubling away while the drive is spinning. In that situation I'd do two things: A) Keep the ontime at a minimum (i.e. rescue most important stuff first, 1st) by file, 2) by dir, 3rd) utilities that work on larger portions of the drive. All of this with the drive mounted to another installation. B) Vary the physical parameters under which you do this; i.e. the drive started failing while it was at cosy operating temperature = cool down before attempting (A) --one of those magic things that sometimes work... - Horst So I wasn't that far off with: 2) Is the situation caused by a software problem (crash) or a hardware problem (HD)? That determines how you want to invest your energy and time. 3) How old is the drive? --the probability of a HD failing is 100% (it's just a matter of time) If it's an old drive, try to rescue what you can (by mounting to another installation). ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
[Eug-lug]Mandrake 8.2 hangup (? in X11/KDE ?)
I had this since I changed from mdk 8.1 (w/o any problems for more than half a year) to mdk 8.2. I start X11/KDE 2.2.2 as normal user, opening a few consoles in KDE (maybe a browser; suddenly certain processes stop working. Even switching out of X to text console doesn't stop this. I start looking at various status information; soon all is frozen -- very few chances to do real-time diagnostics. Shutdown hangs too, very early on, or doesn't even start. 'ps -ef' hangs between X11 and KDE start (!) -see clips below. CPU load goes up to 5.x for no (good) reason. I had that on my original ext2 partition (which got trashed at one crash), and now on a new installation on ext3 (nicer for reboot -but that's not what I expect from a 'nix OS)-: At a first glance I didn't see anything obvious in the logs (but may have overlocked things at a 1st glance) Any tips or comments? .. Horst == ps -ef = UIDPID PPID C STIME TTY TIME CMD ... root 1924 1 0 00:03 vc/5 00:00:00 login -- horxx root 1925 1 0 00:03 vc/6 00:00:00 /sbin/mingetty tty6 root 1928 1920 0 00:03 vc/1 00:00:00 -bash horxx 2018 1924 0 00:03 vc/5 00:00:00 -bash horxx 2044 2018 0 00:03 vc/5 00:00:00 /bin/sh /usr/X11R6/bin/startx horxx 2055 2044 0 00:03 vc/5 00:00:00 xinit /etc/X11/xinit/xinitrc -- -deferglyphs 16 root 2056 2055 4 00:03 ?00:01:04 /etc/X11/X :0 -deferglyphs 16 ==CONSOLE HANGS HERE FOR EVER== == next line would be normally: (different PIDs at diff. boot)== horxx 2235 2229 0 01:06 tty5 00:00:00 /bin/sh /usr/bin/startkde ... == top: = 12:29am up 26 min, 3 users, load average: 4.88, 3.37, 1.90 93 processes: 91 sleeping, 2 running, 0 zombie, 0 stopped CPU states: 2.7% user, 1.5% system, 0.0% nice, 95.6% idle Mem: 255772K av, 246160K used,9612K free, 0K shrd, 46792K buff Swap: 546168K av,3032K used, 543136K free 55928K cached PID USER PRI NI SIZE RSS SHARE STAT %CPU %MEM TIME COMMAND 2056 root 15 0 26344 21M 2548 S 2.1 8.5 1:08 X 2288 horxx 11 0 9304 8680 7316 R 0.9 3.3 0:05 kdeinit 10498 horxx 15 0 1064 1064 812 R 0.7 0.4 0:00 top 2272 horxx 9 0 9268 9268 8416 S 0.1 3.6 0:01 kdeinit 2285 horxx 9 0 9148 9148 8392 S 0.1 3.5 0:00 kdeinit 1 root 8 0 504 504 440 S 0.0 0.1 0:04 init 2 root 9 0 00 0 SW0.0 0.0 0:00 keventd 3 root 9 0 00 0 SW0.0 0.0 0:00 kapmd 4 root 19 19 00 0 SWN 0.0 0.0 0:00 ksoftirqd_CPU0 5 root 9 0 00 0 SW0.0 0.0 0:00 kswapd ... ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]] dma ... // swap Qs -- closing remarks
util-linux 2.9 was released by Andries Brouwer Andries responded right away, and after some back-and-forth, I am attaching the relevant clips. It seems like no one has found a current kernel that supports what is indicated in the man pages of mkdisk regarding cpuinfo, but formally it's OK as stated there (see also his note at the bottom regarding Mandrake's customization). In my opinion, Neil wins the contest: dd if=/dev/hdXX bs=4096 count=1 | strings (increase count as needed) Although this discussion seems only relevant for installations prior to 2.2 kernels, you can trick yourself if you have used older 3rd-party tools like PartitionMagic (e.g. V4.0) on that drive along time ago: In that case even a recent distro installs itself w/o giving you a hint that old style swap is used Horst # On Fri, 3 Jan 2003 [EMAIL PROTECTED] wrote: cat /proc/meminfo Hmm. On the first three Linux machines I try, neither /proc/cpuinfo nor /proc/meminfo give the machine page size. What kernel version and what field in /proc/meminfo works for you? Andries my response: I am looking first at 2.4.18-6 (mandrake). This shows for /proc/meminfo Swap(total) and nrSwapPages --so I conclude 559276032/136542 = 4096 byte/page. (entire output ###below###) snip'ed report for RH 8.0 - Horst # /proc/meminfo 2.4.18-6 (mdk 8.2) # total:used:free: shared: buffers: cached: Mem: 261910528 53567488 2083430400 4870144 23990272 Swap: 5592760320 559276032 MemTotal: 255772 kB MemFree:203460 kB MemShared: 0 kB Buffers: 4756 kB Cached: 23428 kB SwapCached: 0 kB Active: 14616 kB Inactive:28708 kB HighTotal: 0 kB HighFree:0 kB LowTotal: 255772 kB LowFree:203460 kB SwapTotal: 546168 kB SwapFree: 546168 kB NrSwapPages:136542 pages # Linux version 2.4.18-6mdk ([EMAIL PROTECTED]) (gcc version 2.96 2731 (Mandrake Linux 8.2 2.96-0.76mdk)) #1 Fri Mar 15 02:59:08 CET 2002 ~ Date: Fri, 3 Jan 2003 13:52:52 +0100 (MET) From: [EMAIL PROTECTED] Subject: Re: [util-linux] man pages, MKSWAP(8), about swap type; Linux 2.2.4 25 March 1999 NrSwapPages:136542 pages Must be a Mandrake extension. The stock 2.4 kernel has MemTotal: %8lu kB\n MemFree: %8lu kB\n MemShared:%8lu kB\n Buffers: %8lu kB\n Cached: %8lu kB\n SwapCached: %8lu kB\n Active: %8u kB\n Inactive: %8u kB\n HighTotal:%8lu kB\n HighFree: %8lu kB\n LowTotal: %8lu kB\n LowFree: %8lu kB\n SwapTotal:%8lu kB\n SwapFree: %8lu kB\n, in proc_misc.c. So, it is no good to refer to /proc/meminfo. Andries # ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]] dma errors on dd // swap Qs // how to report?
On Wed, 1 Jan 2003, Neil Parker wrote: On Wed, 1 Jan 2003, Linux Rocks ! wrote: my cpuinfo didnt seem to have anything pertinant to page size, however others might... cpuinfo seems a lot less relevent than meminfo... Why would cpuinfo be included, but not meminfo? I dunno. Maybe that part of the man was written long ago when the contents of /proc were different, and never got updated? - Neil Parker Either that, or a typo that kept on propagating (at current state, the difference seems mute // mdk 8.2 shows the same(faulty) man page ). So how do we report this to be corrected ? (I didn't find a ref to the author) BTW - Neil, thanks for clarifying about SWAP_SPACE vs. SWAPSPACE2 in earlier message :-) BTW2 -- after all that confusion I stopped believing anything and tested my system by creating a monstreous object of 500MB on a machine with 256MB RAM and 512 MB swap by using a few python commands to create lists of lists of lists (1KB - 1MB - 500MB); deepcopy in python is slow enough that you can watch the process in a second console using 'top'... quiet entertaining :) , and assuring me that all my swap is implemented as 'new' type. - Horst ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]] dma errors on dd // swap Qs
as Ralph said -- or, since I remember you having several HDs, could use swapoff /dev/hdx* andswapon /dev/hdy* selectively by looking at 'cat /proc/swaps' or 'swapon -s'. Look at mkswap to create a swap partion or file on another drive. Q to all: is the old vs. new swap style issue still relevant these days (old style being limited to 128MB usable space) -- I made a 500 MB swap partition for my 265MB RAM, and top shows it all as avail. -so I'm assuming I config'ed and use the new style ? cat /proc/cpuinfo didn't tell me any more either . Horst On Tue, 31 Dec 2002, Ralph Zeller wrote: swapoff -a On 12/31/02 09am, Dexter Graphic wrote: I just tried copying my partitions one at a time with DMA support turned on and there were no DMA errors reported. You may recall that when I copied the entire drive at once (dd if=/dev/hda of=/dev/hdb bs=1k) with DMA turned on I got DMA and CRC errors in both Debian and Red Hat. So I wonder if this may have been caused by copying my active swap partition? Is there any way to turn off use of the swap portion during my dd backup process? -Dex ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]] dma errors on dd // swap Qs
On Tue, 31 Dec 2002, Linux Rocks ! wrote: cat /proc/cpuinfo should tell you about your CPU. cat /proc/meminfo should tell you about your memory. (including info about swap) Jamie Sounds more than reasonable (and works :-) - I don't know why I blindly trusted the man pages ( RH 8.0 )-: MKSWAP(8) Linux Programmer’s ManualMKSWAP(8) ... If you don’t know the page size that your machine uses, you may be able to look it up with cat /proc/cpuinfo (or you may not ‐ the contents of this file depend on architecture and kernel version). ... Linux 2.2.4 25 March 1999 MKSWAP(8) On Tuesday 31 December 2002 12:27 pm, Horst wrote: ... : Q to all: is the old vs. new swap style issue still relevant these days : (old style being limited to 128MB usable space) -- I made a 500 MB swap : partition for my 265MB RAM, and top shows it all as avail. -so I'm : assuming I config'ed and use the new style ? : cat /proc/cpuinfo didn't tell me any more either . Horst ... ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]Deleted? // topic bind/sendmail grub
Hhm, sounds like interesting material for an open-format LUG topic. I'd love to see that conversation between the two of you 'on stage' at EFN HQ. I have one client where I have to make sendmail/bind configurations, eventually getting it done somehow, never w/o bruises, never with sufficient time to experiment since their site is alive with people waiting for mail delivery... in short, I definitely would be interested. As for the format -casual round-table presentations- I thought our first run in Fall turned out well; then Cory and Ralph set higher standards, but it doesn't have to be always like that all the time, as far as I am concerned. To throw in one more, recently Joseph Crtr. expanded on grub vs. lilo. I'd love to hear more about that at, let's say as a 20 min pres. followed by Q/As -- in connection with another topic, like the one above. My 5 centi-euros Horst On Mon, 23 Dec 2002, Jacob Meuser wrote: On Tue, Dec 24, 2002 at 05:28:30AM +, Bob Crandell wrote: ;; ANSWERS: ocicat.crandell.net. 7146CNAME sd1.mailbank.com. sd1.mailbank.com. 1426A 64.15.175.5 It looks like ocicat and sd1 are the same machine. I don't know how that works. ocicat.crandell.net is 192.168.1.2 The name servers at mailbank.com are saying that ocicat.crandell.net is a CNAME (alias) for sd1.mailbank.com. I set up an alias that looks like: [EMAIL PROTECTED]: [EMAIL PROTECTED] Now it sends but they refuse it. Sigh. ; DiG 2.2 crandell.net.criticalpath.net MX ;; res options: init recurs defnam dnsrch ;; got answer: ;; -HEADER- opcode: QUERY, status: NOERROR, id: 58505 ;; flags: qr rd ra; Ques: 1, Ans: 0, Auth: 1, Addit: 0 ;; QUESTIONS: ;; crandell.net.criticalpath.net, type = MX, class = IN ;; AUTHORITY RECORDS: criticalpath.net. 406 SOA ns1.cp.net. hostmaster.cp.net. ( 2643734399 ; serial 1200; refresh (20 mins) 300 ; retry (5 mins) 2149200 ; expire (24 days 21 hours) 10800 ) ; minimum (3 hours) ;; Total query time: 6 msec ;; FROM: funk to SERVER: default -- 192.168.1.88 ;; WHEN: Mon Dec 23 17:07:09 2002 ;; MSG SIZE sent: 47 rcvd: 117 crandell.net.criticalpath.net has no MX ... no mail routing information for that machine is available. Eventually, it comes back to sender, and the sender refuses. I'll try some more variations and see if I can find a way in. I use an explicit MX record for each host, as well as default for the domain. $ORIGIN . $TTL 21600; 6 hours jakemsr.com IN SOA funk.jakemsr.com. postmaster.funk.jakemsr.com. ( 2001011301 ; serial 3600 ; refresh (1 hour) 1800 ; retry (30 minutes) 604800 ; expire (1 week) 3600 ; minimum (1 hour) ) NS sly.jakemsr.com. NS funk.jakemsr.com. A 64.42.83.115 MX 10 funk.jakemsr.com. $ORIGIN jakemsr.com. funk A 64.42.83.116 MX 10 funk sly A 64.42.83.117 MX 10 funk www A 64.42.83.115 MX 10 funk -- [EMAIL PROTECTED] It furthers the question, Is a 99 cent burger a good deal? ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
[Eug-lug]Movie Thread: Bowling for Columbine
Recently, there has been some interest among Eug-LUG'ers in movies shown at local theaters -- I like to point at Bowling for Columbine, directed by Michael Moore, showing at the Bijou http://www.bijou-cinemas.com# 4:45, 7:05 9:25pm I have enjoyed Moore's documentaries since Roger Me. The subject of the film is certainly real. If that isn't enough, the film was also invited to the 2002 Cannes Film Festival, and received a special award. I will see it tonight (Mon) at the 7:05pm show, and talking about it afterwards at the High Street Cafe. Any Qs? -- respond directly and off the list Horst ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]slooow cable modem connection on 'nix
No, I don't think it's Red Hat specific (as far as 'thinking' can go if you are clueless). As I indicated, with other earlier distros I experienced things similar to some extend, but w/o ever documenting transfer rates. Can you expand on what maybe they trhottle base on uname -r ? means ? - Horst ~ On Wed, 18 Dec 2002, Linux Rocks ! wrote: Yet another reason to not use RedScat. so... do you think att is optimized for win98? maybe they trhottle base on uname -r ? Jamie On Tuesday 17 December 2002 09:57 pm, Horst wrote: : Recently I installed RH 8.0 from the demo CDs and noticed very slow : download rates over cable (for the time being completely ignoring : upload). Though I did a custom install there was little to config since RH : detected and suggested DHCP (except the funny thing(*) below). For : security I picked 'medium', that's iptables with DHCP traffic allowed plus : ssh and http service, which I added. : : Web browsing under RH 8.0/cable is slower than on a good phone line : connection under win98. Similar for plain console operation, like scp -- : putty scp on win98 is about 10 times faster (see clips below for a 70MB : download, server far away). I noticed slower transfer rates with older : 'nix installations before, but those were installed before adding cable : (so I just blamed myself for not tweaking). : In short, this is not distro specific, and doesn't seem to be application : layer dependent (on both GUI and console): always, transfer using win98 : applications is much faster )-: : I also added traceroute for both OS's, though only the 1st hop should be : of interest(if at all), right? : : Any hints? --for either cure or diagnostic ? (I am willing to dig through : long logs of ngrep or ethereal *if* needed) : : - Horst : (*) the funny thing on RH 8.0 is that DHCP client doesn't seem to be able : to get a hostname from ATT, thus using the entire MAC address of eth0 plus : other crap as my host name in the command line prompt (leaving only 50% : for me to type commands... until I manually set hostname) : : THe following records follow... : : = win98 / putty scp === : xyz.sql.zip | 23480 kB | 119.2 kB/s | ETA: 00:06:24 |33% : xyz.sql.zip | 69311 kB | 109.8 kB/s | ETA: 00:00:00 |100% : = about 10 min for 70 MB : ^^ : : === RH 8.0 : xyz.sql.zip 10% |** snip| 7004 KB 1:36:31 ETA : ...Killed by signal 2. : = would have taken 1.5 hrs : : : = win98: : Tracing route to 66.178.136.22 (www.efn.org) over a maximum of 30 hops : : 114 ms 9 ms10 ms 10.139.x.y : 2 8 ms 9 ms 8 ms 12.244.85.1 : 310 ms25 ms10 ms 12.244.64.213 : 466 ms13 ms12 ms 12.244.64.209 : 512 ms18 ms13 ms 12.244.64.205 : 617 ms18 ms18 ms 12.244.72.42 : 717 ms16 ms17 ms 12.123.44.57 : 817 ms21 ms17 ms 12.122.5.157 : 975 ms32 ms34 ms 12.122.2.61 : 1033 ms32 ms37 ms 12.123.13.69 : 1137 ms37 ms41 ms 12.123.221.2 : 1237 ms39 ms41 ms 208.186.87.13 : 1341 ms38 ms40 ms 207.173.114.141 : 1488 ms38 ms75 ms 208.186.21.33 : 1542 ms45 ms42 ms 207.173.114.58 : 1645 ms42 ms43 ms 208.186.20.241 : 1760 ms57 ms61 ms 207.173.115.41 : 18 149 ms93 ms 202 ms 208.186.20.129 : 1957 ms58 ms59 ms 216.190.151.142 : 2062 ms61 ms64 ms 66.178.135.110 : 2168 ms68 ms61 ms 206.96.130.251 : 22 *** Request timed out. : 2364 ms67 ms66 ms 66.178.137.37 : 2467 ms 100 ms77 ms 66.178.136.22 : : Trace complete. : : RH 8.0: === : 1 10.139.x.y 9.869 ms 8.270 ms 9.986 ms : 2 12.244.85.1 9.472 ms 11.891 ms 24.743 ms : 3 12.244.64.213 29.435 ms 9.673 ms 8.349 ms : 4 12.244.64.209 15.013 ms 24.729 ms 9.878 ms : 5 12.244.64.205 11.819 ms 13.417 ms 12.651 ms : 6 12.244.72.42 14.729 ms 23.798 ms 19.512 ms : 7 12.123.44.57 18.581 ms 18.220 ms 14.856 ms : 8 12.122.5.157 27.143 ms 14.280 ms 16.127 ms : 9 12.122.2.61 31.305 ms 32.376 ms 37.096 ms : 10 12.123.13.69 32.957 ms 31.939 ms 35.193 ms : 11 12.123.221.2 37.626 ms 37.957 ms 38.613 ms : 12 208.186.87.13 38.813 ms 39.686 ms 38.005 ms : 13 207.173.114.141 73.291 ms 62.231 ms 53.878 ms : 14 208.186.21.33 41.366 ms 40.519 ms 39.219 ms : 15 207.173.114.58 41.493 ms 53.242 ms 42.126 ms : 16 208.186.20.241 46.523 ms 46.071 ms 44.927 ms : 17 207.173.115.41 54.712 ms 94.296 ms 77.950 ms : 18 208.186.20.129 57.697 ms 59.497 ms 56.909 ms : 19 216.190.151.142 56.408 ms 68.434 ms 57.192 ms : 20
Re: [Eug-lug]slooow cable modem connection on 'nix
On Wed, 18 Dec 2002, Po Petz wrote: On Tue, 17 Dec 2002, Horst wrote: Any hints? --for either cure or diagnostic ? (I am willing to dig through long logs of ngrep or ethereal *if* needed) Wild stabs: What kind of ethernet card are you using? What does ifconfig have to say about your nic? Any RX/TX errors or drops? Have you tried unloading your iptables rules temporarily to see if that's a drag? Indeed, the number of TX errors for just a few Reloads on Mozilla is odd; also there is just my NIC and the modem, and I wouldn't know why there should be so many collisions in such a short time !? I should also mention that my modem 'froze' during the short test (resetting a 'frozen' modem requires power cycle OFF/ON --according to the high-tech support of ATT) I had 'frozen modems' before under 'nix, but only very, very rarely under Win where I can download GBs w/o encountering any problem. The cable modem is an RCA model DCM 235, the NIC a RealTek RTL-8029 (all related kernel messages below) Looks like hardware components have a problems under one OS but work under another, i.e. it's not broken. Now, going through the same routine with all iptable rules flushed I didn't see an increase of errors or collisions, but the load of JPGs still crawls along at 14kbps modem 'speed'... ...hmm, I think I'll sleep over it --Horst RX/TX errors/collisions, and NIC/kernel data follow: [root@athlon root]# ifconfig eth0 eth0 Link encap:Ethernet HWaddr 52:54:05:F7:xy:xy inet addr:12.225.x.y Bcast:255.255.255.255 Mask:255.255.255.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:7684 errors:0 dropped:0 overruns:0 frame:50 TX packets:488 errors:12 dropped:0 overruns:0 carrier:24 collisions:204 txqueuelen:100 RX bytes:622497 (607.9 Kb) TX bytes:61992 (60.5 Kb) Interrupt:10 Base address:0xd000 Mozilla reload page: [root@athlon root]# ifconfig eth0 eth0 Link encap:Ethernet HWaddr 52:54:05:F7:xy:xy inet addr:12.225.x.y Bcast:255.255.255.255 Mask:255.255.255.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:7751 errors:0 dropped:0 overruns:0 frame:54 TX packets:537 errors:23 dropped:0 overruns:0 carrier:46 collisions:392 txqueuelen:100 RX bytes:639009 (624.0 Kb) TX bytes:86780 (84.7 Kb) Interrupt:10 Base address:0xd000 ## RealTek RTL-8029 # ... kernel: NET4: Linux TCP/IP 1.0 for NET4.0 ... kernel: IP Protocols: ICMP, UDP, TCP, IGMP ... nfslock: rpc.statd startup succeeded ... kernel: IP: routing cache hash table of 2048 buckets, 16Kbytes ... rpc.statd[579]: Version 1.0.1 Starting ... kernel: TCP: Hash tables configured (established 16384 bind 32768) ... kernel: Linux IP multicast router 0.06 plus PIM-SM ... kernel: NET4: Unix domain sockets 1.0/SMP for Linux NET4.0. ... kernel: ne2k-pci.c:v1.02 10/19/2000 D. Becker/P. Gortmaker ... kernel: http://www.scyld.com/network/ne2k-pci.html ... kernel: AMD756: dev 10ec:8029, router pirq : 4 get irq : 10 ... kernel: PCI: Found IRQ 10 for device 00:0c.0 ... kernel: PCI: Sharing IRQ 10 with 00:07.4 ... kernel: PCI: Sharing IRQ 10 with 00:08.0 ... kernel: eth0: RealTek RTL-8029 found at 0xd000, IRQ 10, 52:54:05:F7:xy:xy. ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]slooow cable modem connection on 'nix
On Wed, 18 Dec 2002, Jacob Meuser wrote: On Wed, Dec 18, 2002 at 03:14:09AM -0800, Horst wrote: ... kernel: http://www.scyld.com/network/ne2k-pci.html Did you read that? Now I did. Wow, if even the *author* of the kernel module says the following I may have to go NIC shopping: PCI NE2000 clones are a bad idea ... This trend has continued to the PCI bus. A NE2000 design makes little sense here. Certainly, it fits my observations; previous installations all used the same kernel module (but Win doesn't). The NIC worked fine in a local 10Mbps network, but cable modem may require something faster... Thanks man! Horst Mike, also thanks for the tip, but given the gravity of the statement above I may just waste time with kernel and/or IRQ tweaking at each new installation. ... ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]slooow cable modem... end-of-story
Thanks Bob, interesting link suggestion, but it came too late -- I read your message on Mike's computer at Mike's store, and left with an Intel PRO/100 S NIC. Now this is so cool - the JPGs in Mozilla pop up like they would be in cache (but they aren't), and scp is twice(!) as fast as what I had under win98 with the old NIC for my 70MB download test, and w/o a single error or collision: xyz.sql.zip 100% |**| 69311 KB05:31 - Horst (still with some academic Qs below) On Wed, 18 Dec 2002, Bob Miller wrote: Here's somebody on the web who had a very similar problem, with a similar solution. http://msgs.securepoint.com/cgi-bin/get/netfilter-0211/37/1.html I wonder if your cable modem is using a larger MTU than ethernet has, Does MTU mean mess(age) transfer unit ? Hhm, though largely acadimic now, if I had to multiplex(*?*) my end users on an about 1GHz bandwidth, larger packets would make sense ((I assumed multiplexing, or is it truly broadband with each end user having their own ~0.5MHz bandwidth at a fixed frequency? --the latter assuring QoS (including for those users that are not using, thus little overall flexibility)) and is repacketizing your TCP stream along the way. That would allow it to send back-to-back packets out the Ethernet interface, which would be the easiest way to confuse the NIC. But repacketizing IP is evil. Hhm, there may be a reason ATT highly recommends a 10/100 NIC although they *average* delivered bandwidth is much below that... Just for a grin, before you shell out for a new NIC, why not try a different release, such as KNOPPIX? You can download it real fast using Windows. (-: Redhat may have introduced a driver bug. Horst wrote: On Wed, 18 Dec 2002, Jacob Meuser wrote: On Wed, Dec 18, 2002 at 03:14:09AM -0800, Horst wrote: ... kernel: http://www.scyld.com/network/ne2k-pci.html Did you read that? Now I did. Wow, if even the *author* of the kernel module says the following I may have to go NIC shopping: PCI NE2000 clones are a bad idea ... This trend has continued to the PCI bus. A NE2000 design makes little sense here. Certainly, it fits my observations; previous installations all used the same kernel module (but Win doesn't). The NIC worked fine in a local 10Mbps network, but cable modem may require something faster... Thanks man! Horst Mike, also thanks for the tip, but given the gravity of the statement above I may just waste time with kernel and/or IRQ tweaking at each new installation. ... ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug -- Bob Miller Kbob kbobsoft software consulting http://kbobsoft.com [EMAIL PROTECTED] ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]cron
One more comment: avoid runnig other large processes (like backups) while your cron.daily is running (often defaults to 2..4 am). Log rotation and analysis of large log files may interfer with a process that works on the same files. My preference for crontabs is to keep and edit them as regular file(s) and start them with crontab [-u root] crontabFileName (the -u option was suggested on older systems when you su into root as another user... it just became a habbit) ... Horst On Mon, 16 Dec 2002, Rob Hudson wrote: Ah, so I don't need the user to run as in a user specific cron? Make sense. Thanks. On 20021216.2203, Bob Miller said ... Rob Hudson wrote: I'm trying to get a shell script to run every Saturday morning at 4am. Instead of editing /etc/crontab like I usually do, i thought I'd set it up in my user crontab. I ran crontab -e as my user, and added the following: 0 4 * * 6 rob /home/rob/bin/bkup Which runs my backup script. Is there something else I need to do? The timestamps on the files say it didn't run this weekend. Remove the word, rob, unless you have a command called rob in the system's default path. You should have been mailed an error when rob was not found -- is mail set up on this machine? -- Bob Miller Kbob kbobsoft software consulting http://kbobsoft.com [EMAIL PROTECTED] ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
[Eug-lug]slooow cable modem connection on 'nix
Recently I installed RH 8.0 from the demo CDs and noticed very slow download rates over cable (for the time being completely ignoring upload). Though I did a custom install there was little to config since RH detected and suggested DHCP (except the funny thing(*) below). For security I picked 'medium', that's iptables with DHCP traffic allowed plus ssh and http service, which I added. Web browsing under RH 8.0/cable is slower than on a good phone line connection under win98. Similar for plain console operation, like scp -- putty scp on win98 is about 10 times faster (see clips below for a 70MB download, server far away). I noticed slower transfer rates with older 'nix installations before, but those were installed before adding cable (so I just blamed myself for not tweaking). In short, this is not distro specific, and doesn't seem to be application layer dependent (on both GUI and console): always, transfer using win98 applications is much faster )-: I also added traceroute for both OS's, though only the 1st hop should be of interest(if at all), right? Any hints? --for either cure or diagnostic ? (I am willing to dig through long logs of ngrep or ethereal *if* needed) - Horst (*) the funny thing on RH 8.0 is that DHCP client doesn't seem to be able to get a hostname from ATT, thus using the entire MAC address of eth0 plus other crap as my host name in the command line prompt (leaving only 50% for me to type commands... until I manually set hostname) THe following records follow... = win98 / putty scp === xyz.sql.zip | 23480 kB | 119.2 kB/s | ETA: 00:06:24 |33% xyz.sql.zip | 69311 kB | 109.8 kB/s | ETA: 00:00:00 |100% = about 10 min for 70 MB ^^ === RH 8.0 xyz.sql.zip 10% |** snip| 7004 KB 1:36:31 ETA ...Killed by signal 2. = would have taken 1.5 hrs = win98: Tracing route to 66.178.136.22 (www.efn.org) over a maximum of 30 hops 114 ms 9 ms10 ms 10.139.x.y 2 8 ms 9 ms 8 ms 12.244.85.1 310 ms25 ms10 ms 12.244.64.213 466 ms13 ms12 ms 12.244.64.209 512 ms18 ms13 ms 12.244.64.205 617 ms18 ms18 ms 12.244.72.42 717 ms16 ms17 ms 12.123.44.57 817 ms21 ms17 ms 12.122.5.157 975 ms32 ms34 ms 12.122.2.61 1033 ms32 ms37 ms 12.123.13.69 1137 ms37 ms41 ms 12.123.221.2 1237 ms39 ms41 ms 208.186.87.13 1341 ms38 ms40 ms 207.173.114.141 1488 ms38 ms75 ms 208.186.21.33 1542 ms45 ms42 ms 207.173.114.58 1645 ms42 ms43 ms 208.186.20.241 1760 ms57 ms61 ms 207.173.115.41 18 149 ms93 ms 202 ms 208.186.20.129 1957 ms58 ms59 ms 216.190.151.142 2062 ms61 ms64 ms 66.178.135.110 2168 ms68 ms61 ms 206.96.130.251 22 *** Request timed out. 2364 ms67 ms66 ms 66.178.137.37 2467 ms 100 ms77 ms 66.178.136.22 Trace complete. RH 8.0: === 1 10.139.x.y 9.869 ms 8.270 ms 9.986 ms 2 12.244.85.1 9.472 ms 11.891 ms 24.743 ms 3 12.244.64.213 29.435 ms 9.673 ms 8.349 ms 4 12.244.64.209 15.013 ms 24.729 ms 9.878 ms 5 12.244.64.205 11.819 ms 13.417 ms 12.651 ms 6 12.244.72.42 14.729 ms 23.798 ms 19.512 ms 7 12.123.44.57 18.581 ms 18.220 ms 14.856 ms 8 12.122.5.157 27.143 ms 14.280 ms 16.127 ms 9 12.122.2.61 31.305 ms 32.376 ms 37.096 ms 10 12.123.13.69 32.957 ms 31.939 ms 35.193 ms 11 12.123.221.2 37.626 ms 37.957 ms 38.613 ms 12 208.186.87.13 38.813 ms 39.686 ms 38.005 ms 13 207.173.114.141 73.291 ms 62.231 ms 53.878 ms 14 208.186.21.33 41.366 ms 40.519 ms 39.219 ms 15 207.173.114.58 41.493 ms 53.242 ms 42.126 ms 16 208.186.20.241 46.523 ms 46.071 ms 44.927 ms 17 207.173.115.41 54.712 ms 94.296 ms 77.950 ms 18 208.186.20.129 57.697 ms 59.497 ms 56.909 ms 19 216.190.151.142 56.408 ms 68.434 ms 57.192 ms 20 66.178.135.110 61.916 ms 65.592 ms 63.018 ms 21 206.96.130.251 67.022 ms 61.514 ms 102.045 ms 22 * * * 23 66.178.137.37 63.124 ms 64.825 ms 65.743 ms 24 66.178.136.22 64.106 ms 64.196 ms 95.871 ms ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
RE: [Eug-lug]Dan's text based computer system // actual more general.
Ben, thanks for doing this! I thinks I know Dan, and he is a nice guy (maybe twice the age of most of us, in that respect a test case on how well 'we' we are doing in community support, helping folks with a limited budget, and an old PC with limited resources) In more general terms, is there a 'distro', as .tgz, on CD, ...your choice that fits such people's needs --to some extend out of the box? (GUI being a NoNo on a P155) - Horst On Sun, 15 Dec 2002, Dexter Graphic wrote: Ben I will forward your offer to Dan. Maybe he'll let you take the computer home for a few days while you figure it out. If I can get Dan to pay me for my trouble, I'll even shuttle it back and forth; he does not have a car either. Dex From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]On Behalf Of Ben Huot Sent: Sunday, December 15, 2002 10:34 To: [EMAIL PROTECTED] Subject: Re: [Eug-lug]Dan's text based computer system I'd be willing to do it for free as I am a disabled veteran and cannot receive money. I don't know how long it would take me though I am sure I could finish it in a few days. I don't have a car though. Dexter Graphic -- Always trying to present things clearly. ... since someone started cowArt: (__) (__) (__) (\/) ($$) (**) /---\/ /---\/ /---\/ / | 666 || / |=|| / | || * |||| * |||| * |||| Satanic cow This cow is a Yuppie Cow in love (__) (__) (__) (OO) (@@) (xx) /---\/ /---\/ /---\/ / | || / | || / | || * |||| * |||| * |||| Cow who drank Jolt Cow who ate Cow who used Jolt to wash psychadelic mushrooms down psychadelic mushrooms ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]Dexter's dd block size speed test
Dex, nice research ! (-: A few comments: some details of what you found (and others commented on) may depend on the intelligence built into the on-board disk controller (re-mapping bad sectors, buffering I/O, ...) - The default bs=512 matches a 'sector' (floppy disk, MBR) and also should remain for backward compatibility of older custom scripts (or custom scripts *should be* specific regarding defaults, (hmm, myself ...). - in case you are considering writing to the author of 'dd', please add this on my behalf: The reported transferred data should not just be 'n+1 records in/out' but 'n blocks + xyz bytes in/out' -- the prog knows 'xyx' (so that's easy to implement), and such output would provide the user with exact information w/o further doubts) - Horst On Sat, 14 Dec 2002, Dexter Graphic wrote: Purpose: I wanted to find out what effect changing the block size (bs=) option of the dd command would have on partition copy speeds. I also wanted to confirm that the default block size (if no bs option was specified) was indeed 512 bytes as someone had said. Procedure: 1.) I used dd to completely fill up a 75MB partition with data from a larger partition. /dev/hda1 = Debian 3.0 root, ext3 /dev/hda3 = an empty, unformatted partition command used = dd if=/dev/hda1 of=/dev/hda3 (Note: the above command stops copying and aborts once it has used up all the available space on /dev/hda3.) 2.) I then copied this 75MB partition to an identical partition on an identical drive using various block sizes and timing the process. time dd if=/dev/hda3 of=/dev/hdb3 bs=xx Results: no bs= 78s 144584+0 records bs=512 78s 144584+0 records bs=1k 38s 72292+0 records bs=2k 38s 36146+0 records bs=4k 38s 18073+0 records bs=5k 39s 14458+1 records bs=50k 38s 1445+1 records bs=500k 39s 144+1 records bs=512k 39s 144+1 records bs=1M 39s 72+1 records bs=5M 39s 14+1 records bs=10M 39s 7+1 records Conclusions: 1. The default block size if no bs= option is specified is 512 bytes. 2. Any block size larger than the default (512 bytes) will double the copy speed, but using larger block sizes will not result in proportionately greater speed increases. 3. The output of dd shows the number of blocks (records) copied plus (+) the number of partial blocks copied. From the above results I recalled that my hda1 partition was initialized using a 4k block size (which was the default block size in cfdisk). 4. The dd man page should be updated to include this basic operational information (conclusions 1 thru 3) so that people don't have to run their own tests to figure out how to use it. 5. The dd code should *probably* be updated so that 4k is the default block size rather than 512 bytes (since this seems to be the default block size on modern hard disks and it results in doubling of the copy speed. Dexter Graphic ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]Mirroring a drive (all partitions and the MBR)
On Thu, 12 Dec 2002, Dexter Graphic wrote: Any suggestions on how I can mirror (duplicate onto an identical drive) my fancy 3 distribution and 8 partition GNU/Linux system? Would dd if=/dev/hda of=/dev/hdb work? If so, do I have to run it from a boot floppy so that no partitions are mounted? That's what I would try if I had two *identical* devices, except that you may want to increase the block size for speed, from the default of 512 byte (at least on all versions of dd I had) to xk or xM. I am not sure if the block size needs to match the granulariy of the device, i.e. would a partial block been read and written? I just tried: dd if=/dev/zero of=/dev/fd0 # to erase the floppy dd bs=1M count=2 if=/dev/hda of=/dev/fd0 # to look at the partial block Although it said 2+0 record in 1+0 records out I had the entire floppy filled with the correct data (I always wrote the floppy to file after each step and used midnight commander 'mc' in hex mode to look at it) Booting from a floppy would also be my choice since things in /var/ constantly change (either dd may throw up (tar does after several tries on a very busy machine) or fs gets corrupted). Also I don't know how the funny stuff in /proc is handled by dd ? (which you don't need to worry about if the drive is not mounted/active) Good luck, and let us know .. Horst ... ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]Mirroring a drive ... # /proc/*
.. Also I don't know how the funny stuff in /proc is handled by dd ? /proc does not exist on disk -- it's created on the fly by the kernel as you access it. Ja, I could have assumed that 'cuz dd works on the device level, and not on the file system level, e.g. like tar . But Dex, keep in mind, as a warning, if you try to backup your entire installation: tar -cfz proc.tgz /proc not only fails but also garbles up the filename of the broken tar file )-: To handle that you need to look at tar options to ex/include the un/-wanted directories. - Horst ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]Use perl to monitor webserver?
Without having the answer you may want to expand on if 'Restarting the server' means rebooting or web server restart, i.e .../httpd restart. Also, when web server slows down does a local 'lynx IP-of-interface' behave the same? Would a frequent cron job be able to detect, e.g. huge amount of httpd children, or other odd symptons,... and then call the reboot or httpd restart? --there may be symptoms visible before the slowdown reaches the extreme. Certainly an every 15 min cron job that calls a logging script with ... logFile='yourChoice' echo ~~~ $yourChoice date $yourChoice echo ~~~ $yourChoice pstree $yourChoice echo ~~~ $yourChoice ps -Afl $yourChoice echo ~~~ $yourChoice top -b -n 1 $yourChoice echo ~~~ $yourChoice ...etc would help with the diagnostics (but don't let it grow out of control) - Horst On Fri, 13 Dec 2002, Tim Howe wrote: I have a heavily used web server that has been deciding to simply stop serving pages for up to 20 minutes at a time... I have already tried all manner of Apache and OS tweeks to stop this but nothing seems to work. Restarting the server puts everything back on track. What I would like to do, until I find a real fix, is have a Perl program try to connect to that machine on port 80, and if it fails to get a page within, say, 5 seconds, to restart the server. I have found a bunch of modules that will ping and or connect to web servers, but none of them seem to have a good way to time the response. Any suggestions? TimH ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]Use perl to monitor webserver?
... Would a frequent cron job be able to detect, e.g. huge amount of httpd children, or other odd symptons,... and then call the reboot or httpd restart? --there may be symptoms visible before the slowdown reaches the extreme. Certainly an every 15 min cron job that calls a logging script with ... I forgot to mention to look at the other cron jobs cron.daily, etc. included, and start the diagnostics a minute AFTER those. That way you see which processes hog which resources. cron.daily can get quiet busy. Horst ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]Mirroring a drive (all partitions and the MBR)
clip from web: Which block size do you think I should use to improve the speed? Is bigger better? Something is odd on efn -- 2 or 3 of my postings never made it (either out or back in), so I am clipping form web archives. clip of 1st lost: ..., one more on the issue of larger block size: it may not only increase speed but is probably easier on the device if you transfer at/above default block size, rather than stuttering along; also the quality of the (analog, magnetic) data may be better if I/O is performed it the device's natural pace (just a careful guess) I'd try bs=1M ... Horst ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]Use perl to monitor webserver?
Something is odd on efn -- 2 or 3 of my postings never made it (either out or back in), so I am resending. Date: Fri, 13 Dec 2002 13:23:37 -0800 (PST) To: [EMAIL PROTECTED] Subject: Re: [Eug-lug]Use perl to monitor webserver? Without having the answer you may want to expand on if 'Restarting the server' means rebooting or web server restart, i.e .../httpd restart. Also, when web server slows down does a local 'lynx IP-of-interface' behave the same? Would a frequent cron job be able to detect, e.g. huge amount of httpd children, or other odd symptons,... and then call the reboot or httpd restart? --there may be symptoms visible before the slowdown reaches the extreme. Certainly an every 15 min cron job that calls a logging script with ... logFile='yourChoice' echo ~~~ $yourChoice date $yourChoice echo ~~~ $yourChoice pstree $yourChoice echo ~~~ $yourChoice ps -Afl $yourChoice echo ~~~ $yourChoice top -b -n 1 $yourChoice echo ~~~ $yourChoice ...etc would help with the diagnostics (but don't let it grow out of control) - Horst Date: Fri, 13 Dec 2002 13:34:21 -0800 (PST) To: [EMAIL PROTECTED] Subject: Re: [Eug-lug]Use perl to monitor webserver? ... I forgot to mention to look at the other cron jobs cron.daily, etc. included, and start the diagnostics log a minute AFTER those. That way you see which processes hog which resources. cron.daily can get quiet busy. Horst On Fri, 13 Dec 2002, Tim Howe wrote: I have a heavily used web server that has been deciding to simply stop serving pages for up to 20 minutes at a time... I have already tried all manner of Apache and OS tweeks to stop this but nothing seems to work. Restarting the server puts everything back on track. What I would like to do, until I find a real fix, is have a Perl program try to connect to that machine on port 80, and if it fails to get a page within, say, 5 seconds, to restart the server. I have found a bunch of modules that will ping and or connect to web servers, but none of them seem to have a good way to time the response. Any suggestions? TimH ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]Mirroring a drive ... bad sectors
Something is odd on efn -- 2 or 3 of my postings never made it (either out or back in), so I am clipping form web archives. clip from web: Another thing that I am concerned about is how dd handles the occurrence of unusable (marked bad) sectors on the target drive? I would only worry about this if it arrises: check using df, or fdisk, or cfdisk to see current status. And ja, do a fs check after the transfer and see what ends up in lost+found .. Horst ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]Spirited Away -- and 'spirit' afterwards ?
On Thu, 5 Dec 2002, Joseph Carter wrote: It looks like we're going to see the 7:15pm showing. Looks like we should meet then at the Bijou at about 7pm? Several of us are going to be coming ...and micro brew at the High Street Cafe afterwards ? Certainly, I'll do both ... Horst ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]network monitoring tool
On Tue, 19 Nov 2002, Bob Miller wrote: At the firewall presentation, you showed me a tool that summarized network traffic. It was curses-based, updated regularly like top, and had several info screens. What was that tool? I'd like to run it here. Bob, I don't know which tool fits your request, but below is Cory's summary of tool. (most of those are obviously not the answer, but since Cory did such a great job putting that all together someone else may bennefit from having this information posted once) .. Horst Troubleshooting and Admin utilities: - ping - Remember the firewall is configured to not respond to ICMP! When sending out, if you get sendto: Operation not permitted errors. Very likely your ruleset is incorrect. # iptables -P OUTPUT DROP # ping 12.12.12.12 PING 12.12.12.12 (12.12.12.12): 56 data bytes ping: sendto: Operation not permitted ping: wrote 12.12.12.12 64 chars, ret=-1 --- 12.12.12.12 ping statistics --- 2 packets transmitted, 0 packets received, 100% packet loss # iptables -P OUTPUT ACCEPT - tcpdump - Very useful for when it should work, but just doesn't. It's immensely helpful to know some of the filter commands so you can see a subset of the traffic that you are having trouble with. For instance, you have trouble portforwarding to an internal server 10.0.0.10: tcpdump -n -i any ip and host 10.0.0.10 Another scenario, your firewall is a dns caching server for your local lan, but your clients can't seem to get resolved names: tcpdump -n -i any udp and port 53 This is what a successful ping should look through tcpdump like from a client (10.0.0.50) to google.com (216.239.53.101). The firewall has an example public ip address 15.225.140.6: 16:26:06.107110 10.0.0.50 216.239.53.101: icmp: echo request (DF) 16:26:06.107162 15.225.140.6 216.239.53.101: icmp: echo request (DF) 16:26:06.208137 216.239.53.101 15.225.140.6: icmp: echo reply 16:26:06.208182 216.239.53.101 10.0.0.50: icmp: echo reply Other flags: -X prints a hex and ascii dump of each packet -q prints less protocol information -v prints more. q and v can be used usefully together -t does not print a time stamp - ethereal - Ethereal is a graphical program that dumps the network like tcpdump. However there are some great features, most notably, Follow TCP Stream, which combines all of the packets for a particular connection extracts all of their data and displays it in ascii or hex. The conversation is color coded to show who is talking. This is an extremely powerful tool for tracing the communications on an application level (ie seeing the client connect to a webserver and say, GET /index.html HTTP/1.0\nUser-Agent: Linux Rules! and the webserver saying, HTTP/1.1 200 OK - nmap - Network Mapper (nmap) is your friend. You can bet your predators are using it. Use it to scan your firewall and hosts. Read the man page, or at least the major parts. Look at options -sS, -P0, -p, -O and -v. Try nmap --help to see the most commonly used options then read up on those. - lsconntrack - Who's connected? With ipchains one could view masqueraded connections with netstat -M. With full NAT this does not work. Here's how to view the natted connections: cat /proc/net/ip_conntrack Here is a nice script that will parse and nicely display all of that info: http://www.hensema.net/software/lsconntrack/ This is useful in troubleshooting for looking at states. Perhaps a rule was wrong and a client couldn't connect. You fix the rule and you know it is correct, however the client can't connect. Look up the connection tracking and perhaps there is a line in the client that says [UNREPLIED] or something. Display it a again and you'll see the time out field counting down. Wait until this is cleared and then try the client again. Note that I have edited this file to manually remove entries, however I can not recommend this practice in a production environment. - iptraf - Bandwidth and other statistical monitoring Try iptraf -g to go right to general statistics showing the amount of bandwidth on all interfaces. Configure it for kbits/sec or kbytes/sec. Look at packet sizes, traffic distribution by ports and a variety of other things. - How to disable firewall and flush all rules - Your policies may be set to DROP so flushing may effectively disable your network interfaces. iptables -P INPUT ACCEPT iptables -P OUTPUT ACCEPT iptables -P FORWARD ACCEPT iptables -F -t filter iptables -F -t nat - iptables - Of course the most useful tool is the interface to netfilter itself. Read the iptables for good information on syntax and other features. Look at the LOG target. It is good for troubleshooting, accounting or security auditing. Be sure to use the limit match module or you'll fill up your logs quickly! The -L and -v options are the most helpful for trouble shooting. -L to look at your rules, -v will show how many packets matched each rule
Re: [Eug-lug]still dead
Scott, your message doesn't really provide specifics regarding the problem(s) ((I was at the Rocky Horror Show on Thu, so I can only guess)) When I see to reboot and load from the Win95 CDROM, but it didn't take. I would ask: - does the BIOS recognize the CDROM ? - can you boot from a Win95/98 floppy disk containing some CD drivers on it, and config.sys and autoexec.bat installing the driver ? - did the MasterBootRecord on the hard drive get messed up? If so, boot from Win95/98 floppy; change to C: ; issue command 'fdisk /MBR' (w/o 'quotes', also assuming the boot floppy sets the path so A:\[optional DIR\]FDISK.EXE can be reached) -- the option '/MBR' is not really documented by Micro$oft, so how to clearly specify target C: is a bit of a guess. Using Norton/Midnight Commander you'll find some hidden 'documentation' in the binary FDISK.EXE : */MBR and /CMBR cannot both be specified. /MBR only operates on drive 1, use /CMBR for other drives. -You must specify a drive number with /CMBR. Horst -- maybe trying a new installation on Halloween simply isn't a good idea ? ~~ On Sat, 2 Nov 2002, Scott MacWilliams wrote: Hello LUGgers, Thanks again for all your help Thursday night. But... My computer is still not responding to CPR. I followed the steps to reboot and load from the Win95 CDROM, but it didn't take. I'll have to go back and get a more detailed description of what it said, but I haven't had a chance to get to it yet. THanks, Scott MacWilliams ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]Looking for other Linux novices
Cory -- GREAT response ! I am wondering if our web page should have a well visible section along the lines like 'Local Success Stories' to lower the contact-Angst threshold for newbies? Personal statements from people in the community may have more weight than a well written, general article. (of course, it's all in the mail archives, but how many people will go through 1 MB/month of data?) - Horst On Thu, 24 Oct 2002, Cory Petkovsek wrote: Master O Planet, The point of the euglug thursday clinics are supposed to provide you with exactly what you are seeking. Hopefully the environment is not making you feel uncomfortable in asking questions. It is very likely that the questions you ask have been asked by other people, even by those you ask. But there was a time when they did not know such things and its now time for them to share those answers with others. Your questions are important and not out of place. The two tracks are solely for the saturday presentations. Anyone is welcome at either track, regardless of skill. They are merely presentations and not work sessions. The clinics on thursdays are for both beginners and experts. Beginners come often with their system to get help setting up things. Just about two and a half years ago I showed up for the first time at the clinics with my laptop trying to get X configured. X 3.3.6 was the latest, and I couldn't figure out the modelines to get my display working. Another problem I had was trying to get exim to deliver my mail. Upon installing it I had chosen the option no configuration, I understand my mail configuration will be broken until I configure it. For some reason, it just wouldn't deliver mail! One of the attendees was able to help me resolve the problem. Through the lug, I was able to work through problems as they came and eventually help others. Now I've incorporated linux into a large portion of my server room. I'm using it for a firewall, vpn, mail filter, web server, and file and print server. In addition I have a linux server on which I run several perl scripts I've written which replicate tens of thousands of graphic images across servers, monitor my internet line and email me the highlights of the Event Logs from my windows servers. The lug is a great place to gain lots of knowledge quickly. However there are a few things you need to do: 1) Come regularly. You get out of it what you put in to it. 2) Ask questions. Unasked questions are unanswered questions. 3) Watch what experienced people do. When you see someone typing someting on their system, watch! Ask them what they are up to. It's not a secret, otherwise they wouldn't be at the lug. Seek out other novices, that's great. But bring it to the lug. You and another want to setup two clients connecting to the same X server. Bring in your systems and work through your HOWTOs there. When you get stuck, fire off some questions and you'll see your learning bypassing all of the toughest obstacles. The point is to learn as much as you can, not to see how flat your head can be, banging it against the wall, right!? Cory On Thu, Oct 24, 2002 at 04:03:23PM -0700, Master O Planets wrote: I am looking for other Linux beginners who want to build a workgroup. I've been lurking on the edge of the mailing list. In and out of the conversation for two years. I have to admit that I don't know much about Linux because I've been clinging onto Windows. I recently upgraded my video card, and due to some complexities, I almost ran for WinXP, but then there is Palladium and all the crap M$ has pulled before! I contacted the card makers forum and it appears there is a workable solution under Linux. I noticed some talk about a two track system: expert and beginner. I want to see if there are other novices in the group that want to work on specific problems and not feel afraid to ask the dumb questions. ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
[Eug-lug]Re: Tomorrow's meeting
Since I can't come tonight(usually Thu. is fine) here is some feedback: * I like the format suggestion, and the topics suggested by Ralph and Cory. I am not sure if Ralph's email made it from herding to EugLUG ? * Days. We tried before to find a common denominator -- w/o luck. Maybe we alternate days so no one gets excluded all the time. Mo-Th works most the time for me Wkends I am often out in the woods (boating/skiing) My 5c.. Horst On Wed, 9 Oct 2002, Bob Miller wrote: I'd like to propose that we do some organizationalizing at tomorrow's (Thursday's) meeting. The regular meeting will start at 6:30 as usual, and the organizational discussion will start at 7:00. Here's what I have on my mind. Format. I propose that we have a presentation/panel/demo one Thursday night each month, and clinics on all other Thursdays. But I'm open to alternatives. Presentations/panels/demos. What topics would we like, and who will volunteer? (BTW, I've contacted a few of you today, and we have some presenters willing to try it.) Web site. We need a webmaster who will take it over from Rob. (Not me!) Publicity. How can we be more visible? How can we attract new folks? Be there or be disorganized. If you can't make it tomorrow night, please send me comments via email or reply to this message. -- Bob Miller Kbob kbobsoft software consulting http://kbobsoft.com [EMAIL PROTECTED] ___ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]solution to python puzzle
Adding a feature (static data) to a language that doesn't really support
it may alway be a bit of hack (unfortunately, python doesn't have neither
static nor private qualifiers )-:
Wasn't one problem to keep the global namespace clean, and to prevent the
garbage collector from cleaning up the 'static' data of interest? From
other mesages (python list) I concluded Bob's function of interest, using
those 'static' data, relates to module random. So why not adding this all
to module random's namespace?
A) setattr(random, 'pubData', [0]) creates 'semi-static' data, but that's
not interesting because it's very public.
B) a class seems to me always the way to go when both, data and methods,
are tied together. (how would you acces n in your C example if you later
on like to add another function using n ?)
C) the persistance of n[] in Bobs original function(n=[0]) may be not
documented, but seems to be intentional and not a compiler bug as
indicated by random.count.func_defaults below. (that was a question in an
earlier message)
It survived all the abuse I could think of: re-definition, module reload,
garbage collector.
Maybe someone (Sean Reifschneider?) can comment on how safe this really
is.
- Horst
So below is my version of 'ugliness', w/o requiring module __future__.
Most of the code is just for demo purposes - the essential code segments
are mentioned above.
###
import random # module to be extended
import gc # to look at garbage
class Counter: # the standard approach
def __init__(self):
self.__count = 0
def count(self):
self.__count += 1
return self.__count
def test():
print random.count() from test(): , random.count()
random.counter.count() # silently increment
if __name__ == '__main__':
# add semi-static, public data to random's name space:
setattr(random, 'pubData', [0])
# add protected counter to random's name space;
setattr(random, 'counter', Counter())
# add func() to random's name space:
def func(n=[0]): # does not have to be inline
get fancy here with unique start or allow set/reset
n[0] += 1
return n[0]
setattr(random, 'count', func)
def func(): print redefined func() #overwrite previous
func()
print in main - random.count: , random.count()
test()
print - garbage collection - bad items: , gc.collect()
reload(random)
print ... and again: ,
test()
print same for random.counter.count(), random.counter.count()
print \n Now attributes of random.count: \n, dir(random.count)
#print random.count.__doc__
print - random.count.func_defaults: , random.count.func_defaults
###
On Fri, 4 Oct 2002, Bob Miller wrote:
Those of you who were at the clinic last night know that I
was asking for help on a weird limitation of Python.
The problem: Consider the function, foo(), in this C program.
#include stdio.h
int foo()
{
static int n = 0;
return ++n;
}
main()
{
int n1 = foo();
int n2 = foo();
printf(%d %d\n, n1, n2);
return 0;
}
It keeps state around between calls, but does not have extra names in
any nonlocal namespaces.
How would you write a function in Python that does the same?
(Note, I don't want a solution that only returns successive numbers.
I might use this to return successive lines of a file or calculate
successive permutations of a sequence or whatever.)
The solution: For some reason, this apparently simple problem doesn't
have any good solutions (that I'm aware of). Here's the best I can
do.
def foo():
n = [0]
def bar():
n[0] += 1
return n[0]
return bar
foo = foo()
That reuses the single global name, foo. First, foo holds a function
that returns the function we need. Then we set foo to the returned
function. The original function is not accessible, AFAIK.
--
Bob Miller Kbob
kbobsoft software consulting
http://kbobsoft.com [EMAIL PROTECTED]
___
Eug-LUG mailing list
[EMAIL PROTECTED]
http://mailman.efn.org/cgi-bin/listinfo/eug-lug
___
Eug-LUG mailing list
[EMAIL PROTECTED]
http://mailman.efn.org/cgi-bin/listinfo/eug-lug
[Eug-lug]Euglug clinic TODAY plus compiling ?
7:15 Apache configuration demonstration Could we add a little compilation exercise afterwards ? Motivation: Some of the error messages in mod_ssl are not informative enough. I would know how to extent the code (ssl_engine_kernel.c) to improve error logging but experienced problems with dependencies from the http* tree (headers, libraries,...). My approach is to do this in some development directory and then only move the module to /usr/lib/apache/libssl.so This is a production server, and I can't just run a full make/install and wait to see what breaks -- again, the idea is to only toggle the old and new libssl.so (plus httpd restart) to test my changes. Reasonable approach ? Anybody up for this tonight ? .. Horst source: mod_ssl-2.8.10-1.3.26.tar.gz apache_1.3.26.tar.gz On Wed, 18 Sep 2002, Larry Price wrote: start 6:30pm Gather around conference table 7:00 futz with ancient hp hub so we can all plug for ethernet 7:15 Apache configuration demonstration 8:03 laugh when mumble drops soda on their kbd 8:35 first mention of /. 8:53 acceptance of Tux post meeting bofh lart-o-matic as you stumble over the cruft in the street out front ...etc... ___ Eug-lug mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
Re: [Eug-lug]mail strippers -- in python
Mike, if you are not overly fond of regular expression and/or want more control over the output I am happy to share a program I wrote in my early python days. It goes through a(several) textfile(s) and scans for anything that looks like an email address. The first level of reporting shows the result in 4 tables for [unique, duplicate, excluded, fragment] addresses each - unique: most relevant / search not case sensitive, but report preserves case (-: - duplicates: most the time those are of no interest - excluded: e.g. things like '[EMAIL PROTECTED]' from YOUR original mailheader ( this still needs a bit of work) - fragment: anything broken with an @-sign You can list and/or sort-by user, domain, original textline, line-number in original file, ... The second level of reporting shows a diff of two data sets for the 4 different tables above (though only 'unique' makes sense -this needs some work too) The third level of reporting decorates the output to what you want, e.g. to quoted, comma separated values. That's what I use it for once a week to extract a few hundred 'User | Host unknown' from our maillog and feed those into SQL: delete from ... where EMail in ( '[EMAIL PROTECTED]', ...); Maybe that's overkill for what you need; and as I said it's from my early python days (so I won't embarrass myself in public) -but lemme know if interested (python 2.x required) .. Horst. On Wed, 18 Sep 2002, Mike O wrote: Okay, no wisecracks or stupid jokes on that one though I'm sure somebody won't be able to help themselves. Here's what I'm looking for. Anybody know of a program that'll go through all the headers in an email and output all of the email addresses in it to a file or address book? __ Do you Yahoo!? New DSL Internet Access from SBC Yahoo! http://sbc.yahoo.com ___ Eug-lug mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug ___ Eug-lug mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
[Eug-lug]ODD (hacker?) entires in apache error_log
Recently I saw an increase of strange entries in our apache error_log. They look like rejected attacks since none of the listed files/directories exist. Of course, I am wondering how much the attacker sees of those files that DO exist (and therefore, would not produce an error log) ??? (the access log shows nothing related) The STRANGE thing is the LACK of time stamp and requesting URL !?! (the first line in example below shows the format of a regular error) Is that an exploit that looks familiar to any one ? - Horst [Tue Sep 3 14:28:55 2002] [error] [client 204.234.x.y] File does not exist: /home/httpd/html/ sh: /etc/rc.d/init.d/httpsd: No such file or directory sh: /etc/rc.d/init.d/inet: No such file or directory grep: /var/log/httpd/access_log.2: No such file or directory grep: /var/log/httpd/access_log.3: No such file or directory grep: /var/log/httpd/access_log.4: No such file or directory grep: /var/log/httpd/access_log.5: No such file or directory grep: /var/log/httpd/access_log.6: No such file or directory ___ Eug-lug mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
[EUG-LUG:3766] Re: ISO 300-500' of CAT5e
YES, solid wire for fixed installation a) less signal deterioration b) safer connection; with female RJ-45 (CAT-5 wall outlet) the little blades that are supposed to cut through the cable's insulation (using that designated press/hammer tool) may not cut completely because the softer stranded cable gets just squeezed in-between the blades w/o cutting the insulation. ((However, with male RJ-45 it can be harder to crimp 8 solid cables, but there are not many cases where that's needed anyhow)) - Horst On Tue, 20 Aug 2002, Jim Darrough wrote: Use PVC and make sure you glue the joints well if you plan to run it outside. Metal is ok too except it's harder to seal. Cat5e boxed cable comes in 1000 foot boxes usually, and it's almost always solid wire. The only time you are likely to see stranded is for patch cables (flexing service). Sincerely, Jim Darrough Beaker wrote: I had assumed you were wiring a building. If you're making cables, be sure you get stranded CAT-5, which is more expensive. The solid wire won't flex many times before it breaks. Yes - its for wiring a building. I had been wondering though about the difference between stranded solid. I'll be mostly running it up to rooms from the basement plus about 75' needs to be run outside underground through some sort of conduit. I'm planning to use regular exterior electrical conduit unless there's some reason not to. Home Depot is $59 I think. Should be Cat5E. How many feet for $59? 1000'/$59 would be a good deal indeed. Beaker
[EUG-LUG:3750] Re: Fwd: spamming
On Tue, 20 Aug 2002, Bob Crandell wrote: I have a client with a problem similar to this but we think he is being spoofed. How does one stop this? He isn't running an MTA. Would be interesting to see if your client is 'black listed' http://ORDB.org/lookup/ or http://www.abuse.net/relay.html I used a perl script to test the site accused of spamming from another domain. The script comes with documentation and can be searched/downloaded as 'rlytest' -- though I am not sure if that's the latestgreatest, but it worked for me a couple of months ago Horst ((and yes, we were black listed)) Larry Price ([EMAIL PROTECTED]) wrote*: From: Tom Woods [EMAIL PROTECTED] Date: Mon, 19 Aug 2002 12:14:54 -0700 (PDT) To: [EMAIL PROTECTED] Subject: spamming ** This message was sent from the EUGLUG message board. Since ** ** the person who submitted this question may not be on the ** ** mailing list, please reply directly or on the message board: ** ** http://www.euglug.org/board.phtml?id=232 ** my server is being used to spam. What can I do to stop this from happening? I am fairly new to linux. Thanks for the help If your mail server is set up as an open relay (so that anyone can connect to it and send mail) you need to close this fast; take the machine in question off the network before reconfiguring it. Now; we'll need to know what MTA you are using as there are some fairly idiosyncratic mail configuration formats out there. likely ones would be sendmail,qmail,exim,postfix, etc. -- Bob Crandell Assured Computing When you need to be sure. Voice 541-689-9159 FAX 240-371-7237 [EMAIL PROTECTED] www.assuredcomp.com Eugene, Or. 97402
[EUG-LUG:3551] getting serial# from any NS ?
1) How do I get the serial number that an arbitrary NS has cached for a specific domainname ? 2) What is this different if the 'arbitrary' NS is actually the Primary for that domain ? (except that 'cached' wouldn't be the right term?) Thanks, -- Horst
[EUG-LUG:3557] Re: getting serial# from any NS ?
On Mon, Jul 22, 2002 at 07:24:23PM -0700, Horst wrote: 1) How do I get the serial number that an arbitrary NS has cached for a specific domainname ? 2) How is this different if the 'arbitrary' NS is actually the Primary for that domain ? (except that 'cached' wouldn't be the right term?) On Mon, 22 Jul 2002, Jacob Meuser wrote: ... snipped $ dig [ ns.example.com ] example.com SOA ANY [ +nssearch ] Thanks Jacob, this got me off the ground (of course, I tried man dig before posting but that particular man page isn't that helpful); bind 9.2.1 Still, from what I see so far only the NSs with SAO show the details - see 4 blocks below (the 3rd only shows the complaint about 'SOA') - Horst [root@ns Work]# dig ns2.google.com google.com ANY +nssearch SOA ns1.google.com. dns-admin.google.com. 2002072208 7200 1800 1038800 60 from server ns3.google.com in 103 ms. SOA ns1.google.com. dns-admin.google.com. 2002072208 7200 1800 1038800 60 from server ns4.google.com in 54 ms. SOA ns1.google.com. dns-admin.google.com. 2002072208 7200 1800 1038800 60 from server ns1.google.com in 73 ms. SOA ns1.google.com. dns-admin.google.com. 2002072208 7200 1800 1038800 60 from server ns2.google.com in 72 ms. [root@ns Work]# dig ns2.google.com efn.org ANY +nssearch there was nothing, H. [root@ns Work]# dig ns2.google.com efn.org SOA ANY +nssearch ;; Warning, extra type option ;; connection timed out; no servers could be reached [root@ns Work]# [root@ns Work]# dig ns.efn.org efn.org ANY +nssearch SOA efn.org. hostmaster.opn.org. 2002071701 10800 3600 604800 600 from server phloem.uoregon.edu in 390 ms. SOA efn.org. hostmaster.opn.org. 2002071701 10800 3600 604800 600 from server ns.efn.org in 390 ms. SOA efn.org. hostmaster.opn.org. 2002071701 10800 3600 604800 600 from server ns1.efn.org in 99 ms. SOA efn.org. hostmaster.opn.org. 2002071701 10800 3600 604800 600 from server ns2.efn.org in 101 ms. [root@ns Work]#
[EUG-LUG:3496] more 'nix machines at Walmart
It seems like Walmart is really taking on the 'dark site' (there was a short piece on NPR recently about Walmart vs. Microsoft in reference to Lindows). Here comes the clip: Date: Thu, 18 Jul 2002 22:49:16 +0200 (CEST) From: Mandrake Team [EMAIL PROTECTED] Subject: Mandrake Linux Community Newsletter - Issue #51 *** M A N D R A K E L I N U X C O M M U N I T Y N E W S L E T T E R Issue #51 Thursday, 18 July 2002 *** ... Top Story New Mandrake/Microtel systems now at WalMart.com. WalMart.com has released a range of PCs that come pre-loaded with the Mandrake Linux 8.2 operating system. These new Mandrake/Microtel systems are pre-installed with hundreds of the best Open Source applications *plus* the complete StarOffice® 6.0 office suite. Read the complete announcement at MandrakeSoft.com: http://www.mandrakesoft.com/company/press/pr?n=/pr/products/2315 The new systems range from a 900MHz Duron with 128MB RAM for $391, up to a 2.0GHz Pentium 4 with 256MB RAM for $648. See the Desktop PC section of WalMart.com for all the details: http://www.walmart.com/catalog/catalog.gsp?cat=41937dept=3944path=0%3A3944%3A3 951%3A41937 ...
[EUG-LUG:3266] Re: Dexter's boot troubles
Dexter, Let me start by saying that I DO enjoy and appreciate your detailed problem descriptions --and sometimes success report, i.e. how you got the CD burner working. Also my first response was that bits and bytes on your HD are slowly crumbling away. However, if Jamie's remarks regarding unlinked boot block solved your problem my comments below are mute (I wrote them before reading the whole story) Such HD problems can be intermittent and non-reproducible, and may not always be caught by/during checks (that's just my guess). The only way I could identified that once on my HD (and a flaky ZIP drive too) was by making a binary file comparison of the offending files reported, using an earlier backup. For backups I usually tar up entire directory branches or partitions of a working installation I feel is worth keeping, e.g. tar -czf distroX_hdbN_boot_date-plusmore.tgz /boot (since I am too lazy to keep notes I put all I need to know into the file name) There are tons of options documented for the tar command and various ways of doing that, but one warning: you may want to uncompress into a sub-directory for such comparison. How old is your HD ? Is it working hard ? (e.g. swap operations) If the HD is going bad there is no point in wasting time on a detailed analysis. My 5c Horst (on digest). On Tue, 2 Jul 2002, Dexter Graphic wrote: My Mandrake 8.2 system recently stopped booting. After ... Here is some additional information that might be useful. I had done nothing to the system, configuration wise, prior to the boot problem. The last time I used it everything had worked fine and there were no errors or crashes. The computer was off the few days since my last use. It is not a dual boot system, only Mandrake. ... The next day when I turned the damn thing back on, it complained about not having been shut down cleanly and proceeded to do a disk check. It found some errors and I told it to go ahead and fix them--unlinked fragments and stuff like that. The funny thing is that afterwards it booted off the hard disk just fine. I no longer needed to use the boot floppy or CD. My original problem had somehow been solved. Even though I was glad for that, it bugs me terribly that my computer would mysteriously stop working and then equally mysteriously start working again. ... Dexter
[EUG-LUG:3106] Tonight / summer Wednesdays(?) -- programming (python) chat ?
Last Thu. at the weekly meeting I brought up the question if people would be interested in working together on programming projects, preferably python since some of us will be taking the 'Programming for Fun' course (python), or took it in the past. The demand for helping visitors with new installations during the summer months is fairly low, so this would keep some of us focussed and connected during the summer's Wed-meetings (if that's the day chosen) About the format: although there should be something useful resulting from such efforts I am particularly interested in RealLife interaction (discussion and critique) while going through code or working on design issues --the kind of interaction you don't get from mailing lists or by digging through FAQs. I was intentionally vague (and open) regarding potential projects and other issues -just to see if there is interest at all. I'll be at the meeting (EFN, tonight -right?) -- and/or reply either directly or to all, whatever suitable .. Horst.
[EUG-LUG:3032] Re: change of pace
I prefer Wednesdays, because of conflicts with other activities. In addition, people could walk from the TechBrew (also We.) straight to EFN. EFN is the best location so far. In addition to the space, having fast connections and extra terminals available (some with cygwin) makes it possible to address issues other than just installations on the machines people brought in -- I remember the evening Ralph shared with me a number of neat tricks on the console. - Horst, on digest (i.e. 1/2 a day behind) On Wed, 19 Jun 2002, Larry Price wrote: On Wed, 19 Jun 2002, Cory Petkovsek wrote: ...What are the thoughts of a LUG on Tuesday? How does that work for others? I'm not sure where the start of this thread is or who the original author is, but tuesdays are better for me. Thursdays are out completely, which is why I haven't been to the lug in a long while. Cory On Tue, Jun 18, 2002 at 11:44:25AM -0700, Beaker wrote: Most Tuesdays work for me. Unfortunately tuesdays don't work very well for EFN/OPN as that has traditionally been the day for EFN board meetings. I'm going to whack together a PHP page for a sense of the LUG, vote; it'll be available later today.
[EUG-LUG:2834] spec'ing a server -- dual processors ?
I am in the process of consolidating 2 servers[A,B] into 1 machine: [A] does mainly httpd and mysqld; [B] handles sendmail with often fairly large and long jobs. [A] is a 1GHz / 512MB machine, and with 4000 visitors per day (many with DB access) performs quiet well (CPU load is below 0.2 most of the time and rarely above 1.0). [B] is a lame duck (330 MHz), and slows down under load, but that is currently no problem since apache runs on the other machine. Machine [A] has to be replaced with something in the same range (or more MHz/MB Ram ???) and I am contemplating to get a dual processor machine for this. I have little experience with dual processor machines and am wondering how much one can gain in a scenario like that ??? , and what other potential issues I'll have to deal with with dual processors (it will be most likely a RH installation; 7.2?) - Horst
[EUG-LUG:2705] Re: Defunct (zombies)
I experienced zombies on a system that had too many requests (httpd, mysqld, and then analog, i.e the disaster happened at 4 am) for too little memory (the system swapped itself to death). To understand what was going on I ran a cron job at every 15 min that logged: a) date (for time stamp) b) w(for cpu load (1,5,15 min average) c) pstree -l -n (for compact view) d) ps -Afl (for the details) (e) top (?forgot switch for 1 event?) - but top from non-konsole jobs (cron and at) didn't work on older distros(couldn't get default terminal info for non-existing terminal). top also shows zombies in the header. I noticed the zombies are all 'sh' , some with pretty close PID - that's a hint. Good lueck ... Horst. ~~ On Thu, 23 May 2002, Bob Crandell wrote: Hi, I have a server with a growing number of these: 20344 ?Z 0:00 [sh defunct] 20354 ?Z 0:00 [sh defunct] 20355 ?Z 0:00 [sh defunct] 20363 ?Z 0:00 [sh defunct] How do I get rid if them? How do I find out what's causing them? -- Bob Crandell Assured Computing When you need to be sure. Cell 541-914-3985 FAX 240-371-7237 [EMAIL PROTECTED] www.assuredcomp.com Eugene, Or. 97402
[EUG-LUG:2564] Re: How does MandrakeUpdate work?
I haven't seen any response so far, and my answer doesn't exactly address the central questions (auto-upgrade/pay-system/or free) But if you like more control over what will happen on your machine a manual upgrade is described in: http://www.mandrakesecure.net Then click a patch: (it will show you some details) plus the text below Select an FTP server - - download .rpm Since you are tfp'ing from another site you may want to get the GPG key also indicated in the text Finally, a note a bit beyond just a patch upgrade, I have never 'automatically upgraded' to a new Mandrake disto (only reinstalled) since everybody told me it doesn't work as promised -- I would love to hear that's not the case anymore? - Horst Upgrade: To upgrade automatically, use MandrakeUpdate. If you want to upgrade manually, download the updated package(s) from one of our FTP server mirrors and upgrade with rpm -Fvh *.rpm. Verification: Please verify the update prior to upgrading to ensure the integrity of the downloaded package. You can do this with the command: rpm --checksig package.rpm You can get the GPG public key of the Mandrake Linux Security Team to verify the GPG signature of each RPM. If you use MandrakeUpdate, the verification of md5 checksum and GPG signature is performed automatically for you. On Sun, 5 May 2002, Dexter Graphic wrote: I encountered the following instructions on how to install updates to my Mandrake system but it wouldn't work when I tried it. Does anyone know if you have to be on some payment system to get these updates to work? I searched all over their web site trying to find out but I could not find anything describing the operation of MandrakeUpdate. http://www.mandrakesecure.net/en/advisories/2002/MDKA-2002-004-1.php A few packages are being provided as bugfix updates to the Mandrake Linux 8.2 release. ...To upgrade automatically, use MandrakeUpdate. Dexter
[EUG-LUG:2514] Re: LTSP meeting EUGLUG meeting
On Wed, 1 May 2002, Seth Cohn wrote: Yes, the idea of using the Linux meeting as a starting point was my idea. The overlap between the interest is large enough, and if people are coming for the EUGLUG meeting, and aren't interested in the LTSP (and other open source-ish projects on the table to discuss) they can do something else for a bit... I was the one who suggested the date/time, so blame me. Seth ... To be fair, Chris Calise had announced that a long time ago (see ~~~clip~~ w/o all the formatting clutter). He came to a Thu. clinic at Emerald Park about a month ago. I talked a bit with him, and then he talked to Bob (VP) and gave him his card. I thought Chris had some good ideas, and if it wouldn't be for the David Basarmian talk today at the same time I would be at EFN Horst. From: Edward Craig Subject: [EUG-LUG:2282] New Open Source Group Date: Thu, 11 Apr 2002 21:35:19 -0700 This bounced message looks interesting... From: Chris Calise [EMAIL PROTECTED] Greetings, You are cordially invited to an introductory brainstorming session to create a group to promote and support the use of open source software and the Linux Terminal Server Project in Lane County. The mission of this group will be to serve as an umbrella organization for local Linux advocacy groups and individuals volunteering for this project. Through the beginning efforts Tim Rogers, JR Dallas of Head Start of Lane County and myself, we have received a request for LTSP projects at Pleasant Hill Middle school (10 terminals) and also Dexter library. We will need volunteers to help train and support administrators and/or teachers as this project expands. One of the issues on the opening agenda will be how to best implement this support system (mentors?). It is my hope to keep this ball rolling and to expand these pilot LTSP networks to as many schools, libraries, governments, and non-profits as possible but your volunteer help is needed. It is also my hope that this umbrella group will be an effective force for change in Lane County. Where: Eugene Free Net office 43 W. Broadway Eugene When: May 2 7 pm I hope you will be able to attend, and bring your ideas. Thank you. Chris Calise 607-0126 Here are some links to explain this project: http://staff.lanecc.edu/lccoscg/Registry.htm
[EUG-LUG:2424] asynchronous file access: ? use chattr +S ?
On kernel 2.2 and 2.4, both ext2: What happens if a cronjob(root) attempts mv /var/spool/mail/bucket someOtherFilename while sendmail is appending another message to user bucket's inbox at the same time. - Is there a chance to corrupt the file, or the file system? - a quick search points to 'chattr +S fileName' to achieve synchronous behavior -- a more extented search extended the confusion on my end... How does the kernel respond to a second write/move request while the file is written to? a) reject the request, or b) wait and do it later? - because of potential risks I hesitate to simply 'try it' I am more interested in a general clarification than a work-around for the mv command, a) because it is a general issue b) because the script logic depends on the presence of the two files involved. - Horst P.S. This is on a running, remote system where modification of fstab and reboot is not a good option.
[EUG-LUG:2425] Subject: Re: kword crashes before it loads
Ben, as another approach you could (as root) useradd -m dummy passwd dummy (...) become user 'dummy' and install what is interesting to you then as root again compare what trashedUser and dummy have/have not. on the console 'ls -la' , but ideal for comparison is the midnight commander 'mc' -if installed (Options/show hidden files) After having copied those to trashedUser change the owner ship chown trashedUser._maybeTrashedUser_ filename (but check in /home/dummy/ first what the correct group is since some applications may use their group id (though I am not sure for regular user accounts) When you are sure you can simplify the many chown commands by using wildcards for files ( .bash* and so on...) or -R for recursive for entire directories ( .kde/ and so on) Well, this is a nice exercise, but quiet a bit of work if this is all new to you. You may be better off by just creating a new user and copying your old data and documents to the new account. Good luck .. Horst CLIPPED FROM ARCHIVES: = On Thursday 25 April 2002 15:11, Jacob Meuser wrote: On Thu, Apr 25, 2002 at 02:57:18PM -0700, Mark Bigler wrote: That should recreate all the default config files and restore your home files. From there you'll need to reestablish the settings you want. But, I'm sure there are many other cleaner ways to get things back. Why delete the user account and then recreate it? That'll just reinstall .profile/.bashrc, no? Don't those live somewhere in /etc anyway and could just be copied from there?
[EUG-LUG:2336] more Revolution!
I finally got a 2nd VCR to make a copy from that tape (see below)). So, I just returned Bob's and an additional copy I made to the EFN office(Seth), for others to check out ... Horst. Subject: [EUG-LUG:2168] ... / plus was Revolution OS ... Our illustrious VP, Mr. Crandell, has a couple of VHS copies of the movie, Revolution OS. He wants to lend/give them to anybody who wants to watch them. He will have them at this[last] Thursday's clinic.
[EUG-LUG:2279] mysqld and logrotate (or why not?)
PROBLEM:
cron.daily/weekly, logrote creates new /var/log/mysqld.log
but still writes to old /var/log/mysqld.log.N (N = 1,2,...)
I tried all kinds of :
mysqladmin -u root -p refresh
mysqladmin -u root -p flush-logs
mysqladmin -u root -p status
mysqladmin -u root -p flush-logs
mysqladmin -u root -p flush-tables
SOME OBSERVATIONS: (see clips from console)
a) pid of mysqld is only 3byte (LF missing)
b) the pid listed is not the hightest parent
c) logrote instruction for mysql are quiet different from other processes
x)020407 4:02:02 /usr/libexec/mysqld: Binlog closed, cannot RESET
MASTER
'SOLUTION':
The only cure is '.../mysqld restart' - currently run 1hr after
logrotate.
But that's not a 'solution' I am proud of...
Anybody an idea? ... Horst.
MYSQL:
===
[root@plutobin]# cat /etc/logrotate.d/mysqld
/var/log/mysqld.log {
missingok
create 0640 mysql mysql
prerotate
[ -e /var/lock/subsys/mysqld ] /bin/kill -HUP `/bin/cat
/var/run/mysqld/mysqld.pid` || /bin/true
endscript
postrotate
[ -e /var/lock/subsys/mysqld ] /bin/kill -HUP `/bin/cat
/var/run/mysqld/mysqld.pid` || /bin/true
endscript
}
[root@plutobin]#
[root@plutoadmin]# ll /var/run/mysqld/
-rw-rw1 mysqlmysql 3 Apr 5 14:22 mysqld.pid
[root@plutoadmin]# cat /var/run/mysqld/mysqld.pid
582[root@plutoadmin]# ps -ef |grep mysql
root 494 1 0 Apr05 ?00:00:00 /bin/sh
/usr/bin/safe_mysqld --d
mysql 547 494 0 Apr05 ?00:00:07 /usr/libexec/mysqld --
defaults-f
mysql 581 547 0 Apr05 ?00:00:09 /usr/libexec/mysqld --
defaults-f
mysql 582 581 0 Apr05 ?00:00:15 /usr/libexec/mysqld --
defaults-f
mysql 591 581 0 Apr05 ?00:00:00 /usr/libexec/mysqld --
defaults-f
root 20290 19418 0 10:41 pts/100:00:00 grep mysql
OTHER:
===
[root@plutobin]# cat /etc/logrotate.d/named
/var/log/named.log {
missingok
postrotate
/bin/kill -HUP `cat /var/run/named.pid 2 /dev/null` 2 /dev/null
|| tru
e
endscript
}
[root@plutoadmin]# ll /var/run/named/
-rw-r--r--1 namednamed 4 Apr 5 14:22 named.pid
-rw-r--r--1 root root4 Apr 7 04:02 httpd.pid
[root@plutoadmin]# cat /var/run/httpd.pid
930
[root@plutoadmin]# ps -ef |grep httpd
root 930 1 0 Apr05 ?00:00:02 /usr/sbin/httpd -
DHAVE_PROXY -DH
web 18778 930 0 09:24 ?00:00:00 /usr/sbin/httpd -
DHAVE_PROXY -DH
web 18781 930 0 09:24 ?00:00:02 /usr/sbin/httpd -
DHAVE_PROXY -DH
web 18782 930 0 09:24 ?00:00:01 /usr/sbin/httpd -
DHAVE_PROXY -DH
... many more
~~ clip from the mysqld.log* ~~
020407 4:02:02 /usr/libexec/mysqld: Binlog closed, cannot RESET MASTER
Status information:
Current locks:
key_cache status:
blocks used: 2317
not flushed: 0
w_requests: 0
writes: 0
r_requests: 0
reads: 0
handler status:
read_key:0
read_next: 0
read_rnd 0
read_first: 0
write: 0
delete 0
update: 0
Table status:
Opened tables: 0
Open tables:0
Open files: 0
Open streams: 0
020407 4:02:02 /usr/libexec/mysqld: Binlog closed, cannot RESET MASTER
[EUG-LUG:2247] Re: [Fwd: How to get started]
Steve, by now someone has probably already replied to you directly? Just in case -(though I am not the most competent person to reply): YES, you can have Linux AND Win on your machine. If you just like to experiment you can put the Linux boot record onto a floppy disk. That way a) your Win installation doesn't get touched, b) you can't blame Linux if something goes wrong on Win, c) you can freely experiment with different boot constellations because all just goes on different floppies. (e.g. I have a small linux partition on the kid's Win95 machine, and they don't even know about it because I have the linux boot floppy (-: - In order to select HD, partition and boot media you may need to do a customized installation rather than an automated; but that depends on your distro. For starters, Mandrake is easy to install and good in hardware detection. I, or someone else may have a Mdk 8.1 set of CDs. You can get help at our weekly clinics (Thu), but right now I am not sure when we move those from Emerald Park to EFN ??? -check web page or subscribe to list! If you have a second HD designated for linux, or a partition reserved that is not used by Win98 it would be the easiest. If we need to re-sized your current Win98 partition you better get your backup done before you show up at a Thu clinic! 2-3 GB gives you enough room to start with and to have fun. If you have much more space available don't use it all up for one installation --you can have several distros on one HD, and decide later on what you like best, or later on mount that extra space to you favorite installation. Linux is very flexible. I hope that helps ... Horst. On 6 Apr 2002, Rob Hudson wrote: -Forwarded Message- From: Steve [EMAIL PROTECTED] To: [EMAIL PROTECTED] Subject: How to get started Date: 06 Apr 2002 08:39:56 -0800 ** This message was sent from the EUGLUG message board. Since ** ** the person who submitted this question may not be on the ** ** mailing list, please reply directly or on the message board: ** ** http://www.euglug.org/board.phtml?id=180 ** I would like to learn linux, without giving up my windows os. Can i install it on my win98 system without major problems? What type of linux should i start with? Im finding a lot of info on the web, but no basics yet. thanks, steve -- Rob rob_at_euglug_dot_net my @euglugCode = qw(v+++ e--- eug+ bsd+++ gnu+ S+++);
[EUG-LUG:2238] Re: why? could be ipv6 - howto de-ref?
long story short. -rwxr-xr-x6 root root66812 Jan 27 2001 /usr/bin/host -rwxr-xr-x6 root root66812 Jan 27 2001 /usr/bin/ Well, doesn't the 6 indicate there are 4 more surprises to come, like '' but I still wouldn't know how to de-reference what else points to the same I-node ? ... except touching one and making a global find for that exact mtime (pretty goofy) . Horst On Sat, 6 Apr 2002, Larry Price wrote: On Sat, 6 Apr 2002, Jacob Meuser wrote: On Sat, Apr 06, 2002 at 02:16:26AM -0800, Larry Price wrote: which suggests to me that /usr/bin/ is for looking up ipv6 addresses but why the duplication? why not a link? Does work the same when called by a different name? no $ ln /usr/bin/ chunk $ ./chunk efn.org efn.org A 206.163.176.5 so when called as something other than it acts as host normally would. Ah, I see, hard links are not explicitly listed as such by ls -l oops. /me turns pinkish with embarassment -- [EMAIL PROTECTED] http://www.efn.org/~laprice( Community, Cooperation, Consensus http://www.opn.org ( Openness to serendipity, make mistakes http://www.efn.org/~laprice/poems ( but learn from them.(carpe fructus ludi)
[EUG-LUG:2239] Re: EFN grand opening party?
Sorry I missed it -- being on digest... It would be nice to have an Anouncement list for that sort of things. - Horst On Fri, 5 Apr 2002, Seth Cohn wrote: 5:30-8pm tonight 43 West Broadway, efn's new digs. Live Music, party food, and more... I recommend parking in the overpark 1 block north, since it will be free after 5pm... ...
[EUG-LUG:2195] Re: conf sendmail for INBOUND traffic ??? - RH 7.2(!?)
Well, since everybody exept VP (thank Bob !) was busy with HP 9000/300's I had to figure out myself (nice learning experience). To share with the rest, just in case someone else runs into this: This RedHat 7.2 comes with the default DAEMON_OPTIONS(`Port=smtp,Addr=127.0.0.1, Name=MTA') in /etc/mail/sendmail.mc which you have to disable and than reinstall m4 /etc/mail/sendmail.mc /etc/sendmail.cf - Horst. ### ### On Wed, 3 Apr 2002, Horst wrote: Any pointer is appreciated, including additional processes and conf's required/restarted, as well as other 'trivial' and security issues (since I never did this) - I know, there is a rich documentation, but fully digesting that 130KB doc is something I don't look forward to... Thanks much Horst. problem summary: Redhat 7.2, fresh installation. sendmail user1 to user2 on local machine works. sendmail outbound to remote machine works replying(or from scratch) from remote fails (see clip from efn) (all real names are replaced so hostname reads like 'machine1.domean.com' After editing 'local-host-names' executing 'make' in /etc/mail/ looked OK (see *.db below) [root@machine1 mail]# /etc/rc.d/init.d/sendmail restart Shutting down sendmail:[ OK ] Starting sendmail: [ OK ] [root@machine1 mail]# /etc/rc.d/init.d/sendmail status sendmail (pid 11927) is running... [root@machine1 mail]# ~~ who I am: ~~~ [root@machine1 conf]# hostname machine1.domean.com [root@machine1 conf]# errors: # ~~~ recieved at remote account and replied ~~~ ~~ ... but: ~~ The original message was received at Tue, 2 Apr 2002 16:45:43 -0800 (PST) from [EMAIL PROTECTED] [206.163.176.5] - The following addresses had transient non-fatal errors - [EMAIL PROTECTED] - Transcript of session follows - [EMAIL PROTECTED]... Deferred: Connection refused by machine1.domean.com. Warning: message still undelivered after 4 hours Will keep trying until message is 5 days old [ Part 2: Delivery Status ] Reporting-MTA: dns; clavin.efn.org Arrival-Date: Tue, 2 Apr 2002 16:45:43 -0800 (PST) Final-Recipient: RFC822; [EMAIL PROTECTED] Action: delayed Status: 4.4.1 Remote-MTA: DNS; machine1.domean.com Last-Attempt-Date: Tue, 2 Apr 2002 20:57:10 -0800 (PST) Will-Retry-Until: Sun, 7 Apr 2002 17:45:43 -0800 (PST) ~~~ another error, but local: ~~~ less /var/log/maillog ... Apr 3 09:45:26 localhost sendmail[11367]: NOQUEUE: machine1.domean.com [127.0.0.1] did not issue MAIL/EXPN/VRFY/ETRN during connection to MTA ... ### directories, files, file content: [root@machine1 mail]# pwd /etc/mail [root@machine1 mail]# ls -ltr total 68 -rw-r--r--1 root root0 Mar 2 2001 virtusertable -rw-r--r--1 root root 127 Mar 2 2001 trusted-users -rw-r--r--1 root root 2220 Mar 2 2001 sendmail.mc -rw-r--r--1 root root 611 Mar 2 2001 Makefile -rw-r--r--1 root root0 Mar 2 2001 mailertable -rw-r--r--1 root root 5573 Mar 2 2001 helpfile -rw-r--r--1 root root0 Mar 2 2001 domaintable -rw-r--r--1 root root 155 Apr 2 15:02 local-host-names -rw-r--r--1 root root 528 Apr 2 15:54 access -rw-r--r--1 root root12288 Apr 2 16:06 virtusertable.db -rw-r--r--1 root root12288 Apr 2 16:06 mailertable.db -rw-r--r--1 root root12288 Apr 2 16:06 domaintable.db -rw-r--r--1 root root12288 Apr 2 16:06 access.db drwx--2 root root 4096 Apr 3 09:52 MISC [root@machine1 mail]# [root@machine1 mail]# ls -l /etc/send* -rw-r--r--1 root root46245 Mar 2 2001 /etc/sendmail.cf [root@machine1 mail]# grep '/etc/mail/' ../sendmail.cf Fw/etc/mail/local-host-names FR-o /etc/mail/relay-domains Kmailertable hash -o /etc/mail/mailertable Kvirtuser hash -o /etc/mail/virtusertable Kaccess hash /etc/mail/access #O ErrorHeader=/etc/mail/error-header O HelpFile=/etc/mail/helpfile O UserDatabaseSpec=/etc/mail/userdb.db #O ServiceSwitchFile=/etc/mail/service.switch #O DefaultAuthInfo=/etc/mail/default-auth-info Ft/etc/mail/trusted-users [root@machine1 mail]# [root@machine1 mail]# cat local-host-names # local-host-names - include all aliases for your machine here. machine1.domean.com [root@machine1 mail]# cat access # Check the /usr/share/doc/sendmail-8.11.2/README.cf file for a description # of the format of this file. (search
[EUG-LUG:2176] conf sendmail for INBOUND traffic ???
Any pointer is appreciated, including additional processes and conf's required/restarted, as well as other 'trivial' and security issues (since I never did this) - I know, there is a rich documentation, but fully digesting that 130KB doc is something I don't look forward to... Thanks much Horst. problem summary: Redhat 7.2, fresh installation. sendmail user1 to user2 on local machine works. sendmail outbound to remote machine works replying(or from scratch) from remote fails (see clip from efn) (all real names are replaced so hostname reads like 'machine1.domean.com' After editing 'local-host-names' executing 'make' in /etc/mail/ looked OK (see *.db below) [root@machine1 mail]# /etc/rc.d/init.d/sendmail restart Shutting down sendmail:[ OK ] Starting sendmail: [ OK ] [root@machine1 mail]# /etc/rc.d/init.d/sendmail status sendmail (pid 11927) is running... [root@machine1 mail]# ~~ who I am: ~~~ [root@machine1 conf]# hostname machine1.domean.com [root@machine1 conf]# errors: # ~~~ recieved at remote account and replied ~~~ ~~ ... but: ~~ The original message was received at Tue, 2 Apr 2002 16:45:43 -0800 (PST) from [EMAIL PROTECTED] [206.163.176.5] - The following addresses had transient non-fatal errors - [EMAIL PROTECTED] - Transcript of session follows - [EMAIL PROTECTED]... Deferred: Connection refused by machine1.domean.com. Warning: message still undelivered after 4 hours Will keep trying until message is 5 days old [ Part 2: Delivery Status ] Reporting-MTA: dns; clavin.efn.org Arrival-Date: Tue, 2 Apr 2002 16:45:43 -0800 (PST) Final-Recipient: RFC822; [EMAIL PROTECTED] Action: delayed Status: 4.4.1 Remote-MTA: DNS; machine1.domean.com Last-Attempt-Date: Tue, 2 Apr 2002 20:57:10 -0800 (PST) Will-Retry-Until: Sun, 7 Apr 2002 17:45:43 -0800 (PST) ~~~ another error, but local: ~~~ less /var/log/maillog ... Apr 3 09:45:26 localhost sendmail[11367]: NOQUEUE: machine1.domean.com [127.0.0.1] did not issue MAIL/EXPN/VRFY/ETRN during connection to MTA ... ### directories, files, file content: [root@machine1 mail]# pwd /etc/mail [root@machine1 mail]# ls -ltr total 68 -rw-r--r--1 root root0 Mar 2 2001 virtusertable -rw-r--r--1 root root 127 Mar 2 2001 trusted-users -rw-r--r--1 root root 2220 Mar 2 2001 sendmail.mc -rw-r--r--1 root root 611 Mar 2 2001 Makefile -rw-r--r--1 root root0 Mar 2 2001 mailertable -rw-r--r--1 root root 5573 Mar 2 2001 helpfile -rw-r--r--1 root root0 Mar 2 2001 domaintable -rw-r--r--1 root root 155 Apr 2 15:02 local-host-names -rw-r--r--1 root root 528 Apr 2 15:54 access -rw-r--r--1 root root12288 Apr 2 16:06 virtusertable.db -rw-r--r--1 root root12288 Apr 2 16:06 mailertable.db -rw-r--r--1 root root12288 Apr 2 16:06 domaintable.db -rw-r--r--1 root root12288 Apr 2 16:06 access.db drwx--2 root root 4096 Apr 3 09:52 MISC [root@machine1 mail]# [root@machine1 mail]# ls -l /etc/send* -rw-r--r--1 root root46245 Mar 2 2001 /etc/sendmail.cf [root@machine1 mail]# grep '/etc/mail/' ../sendmail.cf Fw/etc/mail/local-host-names FR-o /etc/mail/relay-domains Kmailertable hash -o /etc/mail/mailertable Kvirtuser hash -o /etc/mail/virtusertable Kaccess hash /etc/mail/access #O ErrorHeader=/etc/mail/error-header O HelpFile=/etc/mail/helpfile O UserDatabaseSpec=/etc/mail/userdb.db #O ServiceSwitchFile=/etc/mail/service.switch #O DefaultAuthInfo=/etc/mail/default-auth-info Ft/etc/mail/trusted-users [root@machine1 mail]# [root@machine1 mail]# cat local-host-names # local-host-names - include all aliases for your machine here. machine1.domean.com [root@machine1 mail]# cat access # Check the /usr/share/doc/sendmail-8.11.2/README.cf file for a description # of the format of this file. (search for access_db in that file) # The /usr/share/doc/sendmail-8.11.2/README.cf is part of the sendmail-doc # package. # # by default we allow relaying from localhost... localhost.localdomain RELAY localhost RELAY 127.0.0.1 RELAY # tried a few things, but now commented out. [root@machine1 mail]#
[EUG-LUG:2168] ZIP 100 disks for sale / plus was Revolution OS
Mr. VP I'd like a copy; will reimburse you, of course. ZIP 100 disks: I have 10+ disks for sale, only used a few times for backups, recently formated (full, w/o bad sectors); price $5 ea, or less for more, or make an offer. Can bring them on Thu. Please reply directly, w/o cc to list .. Horst Our illustrious VP, Mr. Crandell, has a couple of VHS copies of the movie, Revolution OS. He wants to lend/give them to anybody who wants to watch them. He will have them at this Thursday's clinic.
