Re: Skipping F1 FreeBSD prompt on boot
Pieter de Goeje wrote: On Tuesday 15 May 2007, Sam Lawrance wrote: On 14/05/2007, at 10:41 AM, Pieter de Goeje wrote: On Sunday 13 May 2007, David Landgren wrote: Sam Lawrance wrote: On 13/05/2007, at 6:15 PM, Matthew Seaman wrote: [...] the drive, and likely to remain that way until the disk dies of mechanical failure. I just don't need that prompt, especially the annoying beep it makes. The beep was removed since May 2006 (6.2-RELEASE, 6-STABLE, HEAD). A simple #boot0cfg -B /dev/adX should get rid of it. I thought I remembered that! Wasn't it removed to reclaim a couple extra bytes? :-) Quote from the commit log: Restore the pre-5.x behavior of only beeping if the user makes a bad selection and not always beeping on startup. The two bytes for the extra 'jmp' instruction were obtained by removing recognition of BSD/OS partitions. Cheers, Pieter de Goeje Heh, ok, for extra bonus points, what/where is the code that makes the two annoying BEEPs on shutdown? If I could compile that out, my life would be complete :) Thanks, David ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Skipping F1 FreeBSD prompt on boot
Pieter de Goeje wrote: On Thursday 17 May 2007, David Landgren wrote: Heh, ok, for extra bonus points, what/where is the code that makes the two annoying BEEPs on shutdown? If I could compile that out, my life would be complete :) Thanks, David Hmm, I've never heard any beeps on shutdown... how do you shutdown your system? When I type 'halt -p' it just powers off after synching the disks, no beep whatsoever. shutdown -p now ... so that would mean it's shutdown that does that? The annoyance factor has never been enough to make me investigate more closely. But hey, if halt -p is safe and clean, and silent, that's good enough for me. Thanks, David ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Skipping F1 FreeBSD prompt on boot
Sam Lawrance wrote: On 13/05/2007, at 6:15 PM, Matthew Seaman wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA256 Sam Lawrance wrote: On 12/05/2007, at 8:59 PM, Matthew Seaman wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA256 David Landgren wrote: I have a disk that has only FreeBSD on it, and so I would like to skip the initial F1/FreeBSD prompt. boot0cfg -v ad0 says: options=nopacket,update,nosetdrv default_selection=F1 (Slice 1) ... what do I have to do to say JFDI instead of prompting? This is not the sort of thing I want to fiddle around experimenting, so a little guidance would be most appreciated. fdisk -B -b /boot/mbr /dev/ad0 You installed the FreeBSD boot sector stuff, which gives you the 'press F1' business. Replace that with the standard mbr, which just boots straight up. Rather than replacing it, you can use boot0cfg to set a really short timeout instead; in case you might want that functionality one day. Heh. It's not like you only get one chance to rewrite the boot blocks on any particular drive. If anyone needs to (re-)install the FreeBSD boot blocks, then you can do very simply it by: boot0cfg -B -b /boot/boot0 /dev/ad0 or even fdisk -B -b /boot/boot0 /dev/ad0 Or if you need to boot from a serial console you can change /boot/boot0 to /boot/boot0sio Sure, but why get rid of it, when leaving it in with a short timeout costs you nothing. A fair point, but in this particular case, FreeBSD is the only thing on the drive, and likely to remain that way until the disk dies of mechanical failure. I just don't need that prompt, especially the annoying beep it makes. David ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Skipping F1 FreeBSD prompt on boot
List, I have a disk that has only FreeBSD on it, and so I would like to skip the initial F1/FreeBSD prompt. boot0cfg -v ad0 says: options=nopacket,update,nosetdrv default_selection=F1 (Slice 1) ... what do I have to do to say JFDI instead of prompting? This is not the sort of thing I want to fiddle around experimenting, so a little guidance would be most appreciated. Thanks, David ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Problème avec le floppy disk.
Tribal, Grégory wrote: Bonjour, Nous avons un problème au moment de l'installation d'un de nos logiciel: Lorsque que nous insérons une disquette et lançons notre commande d'installation, le message suivant apparait: /dev/fd0: Cannot read :Input/Output error At Beginning of tape - Quitting now Error is not recoverable: Exiting now Néanmoins, le lecteur de disquette fonctionne car nous utilisons plusieurs autres disquettes avant la disquette d'installation. Nous avons vérifié le BIOS et tout parait bien configuré... Est-ce la disquette fonctionne? Est-ce que le filesystem de la disquette est de type msdosfs? Est-ce que vous pouvez lire la disquette sur une autre machine? Si vous pouvez répondre en anglais, vous aurez plus d'aide, cette liste est majoritairement anglophone. Je ne connais pas de liste francophone existe pour freebsd. Merci, David ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: sparc64 and perl 5.8.8 port test failures?
Steven D. Yee wrote: I'm seeing multiple errors from make test and I can't seem to figure out how to get rid of them. As far as I can tell it builds correctly. [...] ../lib/integer..NOK 10 # Failed test 'left shift' # in ../lib/integer.t at line 49. # got: '-4292583424' # expected: '-9223372036854775808' # Looks like you failed 1 test of 11. ../lib/integer..dubious Test returned status 1 (wstat 256, 0x100) DIED. FAILED test 10 Failed 1/11 tests, 90.91% okay I did try building with WITHOUT_PERL_64BITINT=yes but that didn't seem to make a difference, although its possible that I screwed that up since conf.sh still shows multiple references to 64 bit ints (even use64bitint is defined) does anyone have any pointers as to what may be going on? or where to start looking? This looks like the build is bringing in 64bitness when it shouldn't (or vice versa). The build process might have remnants of the previous config run lying around (in Policy.sh and/or config.sh). Step down into the build directory and delete these two files, and build again. Or, better yet, just delete the entire ./work directory, and build it again. Later, DAvid ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: sparc64 and perl 5.8.8 port test failures?
Steven D. Yee wrote: I'm seeing multiple errors from make test and I can't seem to figure out how to get rid of them. As far as I can tell it builds correctly. [...] ../lib/integer..NOK 10 # Failed test 'left shift' # in ../lib/integer.t at line 49. # got: '-4292583424' # expected: '-9223372036854775808' # Looks like you failed 1 test of 11. ../lib/integer..dubious Test returned status 1 (wstat 256, 0x100) DIED. FAILED test 10 Failed 1/11 tests, 90.91% okay I did try building with WITHOUT_PERL_64BITINT=yes but that didn't seem to make a difference, although its possible that I screwed that up since conf.sh still shows multiple references to 64 bit ints (even use64bitint is defined) does anyone have any pointers as to what may be going on? or where to start looking? This looks like the build is bringing in 64bitness when it shouldn't (or vice versa). The build process might have remnants of the previous config run lying around (in Policy.sh and/or config.sh). Step down into the build directory and delete these two files, and build again. Or, better yet, just delete the entire ./work directory, and build it again. Later, DAvid ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
cyrus imap: seems Ok, but can't connect with mutt on localhost
List, I have recently set up cyrus IMAP (from ports) and it seems to be behaving itself: % imtest localhost S: * OK [CAPABILITY IMAP4 IMAP4rev1 LITERAL+ ID AUTH=DIGEST-MD5 SASL-IR] example.com Cyrus IMAP4 v2.3.7 server ready C: C01 CAPABILITY S: * CAPABILITY IMAP4 IMAP4rev1 LITERAL+ ID AUTH=DIGEST-MD5 SASL-IR ACL RIGHTS=kxte QUOTA MAILBOX-REFERRALS NAMESPACE UIDPLUS NO_ATOMIC_RENAME UNSELECT CHILDREN MULTIAPPEND BINARY SORT SORT=MODSEQ THREAD=ORDEREDSUBJECT THREAD=REFERENCES ANNOTATEMORE CATENATE CONDSTORE IDLE LISTEXT LIST-SUBSCRIBED X-NETSCAPE URLAUTH S: C01 OK Completed C: A01 AUTHENTICATE DIGEST-MD5 S: + bm9uY2U...cw== Please enter your password: C: dXNlcm5h...mZjE= S: + cnNwYXV0aD1hYmQ4ZWIyNmYwOWRlMzEyMjQxNzY4M2ZmN2MxMzI1Ng== C: S: A01 OK Success (privacy protection) Authenticated. Security strength factor: 128 . logout * BYE LOGOUT received . OK Completed Connection closed. Similarly, I can send an SMTP message to port 25, and watch Postfix correctly deliver it to cyrus, which squirrels it away in its spool. As a final test, I thought I should be able to access the IMAP folder on localhost via mutt. But it doesn't seem to work. When I fire up mutt (compiled from ports) and try to change folders: c imap://localhost/INBOX It just hangs, with a Connecting to localhost... status message. All I see in /var/log/debug.log is: Dec 5 23:27:20 localhost imap[56129]: accepted connection Nothing in any other log files. Is there something really obvious that I'm overlooking? Thanks, David -- It's overkill of course, but you can never have too much overkill. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Connection refusal for an NFS mount
David Kelly wrote: On Thu, Jul 20, 2006 at 07:43:16PM +0200, David Landgren wrote: David Kelly wrote: For starters try showmount -e the.freebsd.ip.address on the Linux box to see if the Linux box sees the NFS daemons on the FreeBSD machine. Hrm. # showmount -e 172.17.0.21 mount clntudp_create: RPC: Port mapper failure - RPC: Unable to receive I don't think NFS is going to work until you can get past the above problem. Running showmount -e on your FreeBSD machine should display the essential contents of /etc/exports. I added /var 172.17.0.21 /usr 127.0.0.1 to /etc/exports on the FreeBSD machine, hupped mountd, and when I run showmount -e 172.17.0.21 showmount -e 127.0.0.1 ... either command just hangs indefinitely. Hmm. What does the FreeBSD machine have to say about your attempts to connect from Linux in /var/log/messages? Nothing. Which is reasonable, given the above. So it looks like NFS is hosed on this box. Let's see now, relevant lines from rc.conf firewall_enable=YES firewall_type=open kern_securelevel_enable=NO nfs_access_cache=2 nfs_bufpackets= nfs_reserved_port_only=NO nfs_server_enable=YES nfs_server_flags=-u -t -n 4 mountd_enable=YES ntpd_enable=YES rpc_lockd_enable=NO rpcbind_enable=YES rpc_statd_enable=YES Hmm. I don't what nfs_bufpackets does. Short of rebooting the server, how do I reinitialise the NFS layers? Does the following order sound sane? /etc/rc.d/mountd stop /etc/rc.d/nfsd stop /etc/rc.d/rpcbind stop ... and the the same again with start in the reverse order? Thanks, David -- Much of the propaganda that passes for news in our own society is given to immobilising and pacifying people and diverting them from the idea that they can confront power. -- John Pilger ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Connection refusal for an NFS mount
List, On an old Redhat box (address 172.17.0.18), I'm trying to mount an NFS export from a FreeBSD (5.2.1-RELEASE) box. Both machines are on the same network segment, and neither have any onboard firewalling rules. I run the following command (on the redhat, bechet is the FreeBSD box): mount -v -t nfs -s -o ro,soft,intr bechet:/home/ftp/pub/mirror /net/mir And I get back: mount: RPC: Remote system error - Connection refused On the FreeBSD box, /etc/exports contains /home/ftp/pub/mirror 172.17.0.18 rpcinfo -p produces: program vers proto port service 104 local111 rpcbind 103 local111 rpcbind 102 local111 rpcbind /etc/hosts.deny is empty, and /etc/hosts.allow contains ALL: ALL : allow The relevant processes on the FreeBSD box are 95 ?? IL 0:00.00 (nfsiod 0) 96 ?? IL 0:00.00 (nfsiod 1) 97 ?? IL 0:00.00 (nfsiod 2) 98 ?? IL 0:00.00 (nfsiod 3) 343 ?? Is 0:04.41 /usr/sbin/rpcbind 425 ?? Is 0:04.61 /usr/sbin/rpc.statd 11373 ?? Is 0:00.01 /usr/sbin/mountd -r 88497 ?? Is 0:00.05 nfsd: master (nfsd) 88498 ?? I 0:00.00 nfsd: server (nfsd) 88499 ?? I 0:00.00 nfsd: server (nfsd) 88500 ?? I 0:00.00 nfsd: server (nfsd) 88501 ?? I 0:00.00 nfsd: server (nfsd) (I understand, from reading the handbook, that I should be using rpcbind rather than portmap). This server has been an NFS server in the past, so I know it worked at some point. I'm not sure if I'm missing a daemon in the mix, or if there's something else I've overlooked. Any clues will be most graciously received :) Thanks, David -- Much of the propaganda that passes for news in our own society is given to immobilising and pacifying people and diverting them from the idea that they can confront power. -- John Pilger ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Connection refusal for an NFS mount
David Kelly wrote: On Thu, Jul 20, 2006 at 05:16:16PM +0200, David Landgren wrote: List, On an old Redhat box (address 172.17.0.18), I'm trying to mount an NFS export from a FreeBSD (5.2.1-RELEASE) box. Both machines are on the same network segment, and neither have any onboard firewalling rules. [...] (I understand, from reading the handbook, that I should be using rpcbind rather than portmap). This server has been an NFS server in the past, so I know it worked at some point. I'm not sure if I'm missing a daemon in the mix, or if there's something else I've overlooked. Any clues will be most graciously received :) For starters try showmount -e the.freebsd.ip.address on the Linux box to see if the Linux box sees the NFS daemons on the FreeBSD machine. Hrm. # showmount -e 172.17.0.21 mount clntudp_create: RPC: Port mapper failure - RPC: Unable to receive mountd needs to be running on the FreeBSD host (apparently yours is running). When /etc/exports changes mountd needs to be informed: kill -s HUP `cat /var/run/mountd.pid` Yup, know about that. Also at least in the past Linux distributions defaulted NFS to non-reserved ports. Your Linux may not be talking to the same ports as the FreeBSD machine is listening. Let's have a look... # nmap 172.17.0.21 Starting nmap V. 3.00 ( www.insecure.org/nmap/ ) Interesting ports on bechet.bpinet.com (172.17.0.21): (The 1584 ports scanned but not shown below are in state: closed) Port State Service 21/tcp openftp 22/tcp openssh 25/tcp opensmtp 37/tcp opentime 80/tcp openhttp 199/tcpopensmux 443/tcpopenhttps 801/tcpopendevice 901/tcpopensamba-swat 1011/tcp openunknown 1020/tcp openunknown 2049/tcp opennfs 3306/tcp openmysql 5308/tcp opencfengine 5432/tcp openpostgres 5999/tcp openncd-conf 8080/tcp openhttp-proxy My god there's a lot of crap on that box! Still, looks like NFS is running. And according to the man page of the linux box: port=n The numeric value of the port to connect to the NFS server on. If the port number is 0 (the default) then query the remote host's portmapper for the port number to use. If the remote hostâs NFS daemon is not regis- tered with its portmapper, the standard NFS port number 2049 is used instead. So that sounds about right. I tried adding port=2049 explictly to the mount command, but same error: Connection refused Well, thanks for your help. Beats me what I've done wrong. Thanks, David -- Much of the propaganda that passes for news in our own society is given to immobilising and pacifying people and diverting them from the idea that they can confront power. -- John Pilger ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: php_XML_ParserCreate error
Mike Sacauskis wrote: Hi I'm trying to install mambo on a 5.4 release. I'm running into a problem with the mambo editor. I've traced it to an error in the apache log: /libexec/ld-elf.so.1: /usr/local/lib/php/20020429/xml.so: Undefined symbol php_XML_ParserCreate Hello Mike, I have just run into this same problem. Did you ever get the problem resolved to your satisfaction? I've just rebuilt php using a more recent version (4.4.0), and I'm getting the same error, and I can't see what's missing either. Best, David I did an nm on the xml.so library and the symblols for php_XML_ParserCreate are undefined. 6940 t parserInit U php_XML_ErrorString U php_XML_ExpatVersion U php_XML_GetCurrentByteIndex U php_XML_GetCurrentColumnNumber U php_XML_GetCurrentLineNumber U php_XML_GetErrorCode U php_XML_Parse U php_XML_ParserCreate U php_XML_ParserCreateNS U php_XML_ParserFree U php_XML_SetCharacterDataHandler U php_XML_SetDefaultHandler U php_XML_SetElementHandler U php_XML_SetEndNamespaceDeclHandler U php_XML_SetExternalEntityRefHandler U php_XML_SetNotationDeclHandler U php_XML_SetProcessingInstructionHandler U php_XML_SetStartNamespaceDeclHandler U php_XML_SetUnparsedEntityDeclHandler U php_XML_SetUserData U php_error_docref0 There are macros in /usr/ports/lang/php4/work/php-4.4.1/main that map the php_XML routines to XML_ routines. Is this how these symbols are supposed to be resolved? If so is there something that needs to be configured to allow this? These are the ports I have installed: apache-1.3.34_3 The extremely popular Apache http server. Very fast, very php4-4.4.1_3PHP Scripting Language (Apache Module and CLI) php4-gd-4.4.1_3 The gd shared extension for php php4-mysql-4.4.1_3 The mysql shared extension for php php4-pcre-4.4.1_3 The pcre shared extension for php php4-session-4.4.1_3 The session shared extension for php php4-xml-4.4.1_3The xml shared extension for php php4-zlib-4.4.1_3 The zlib shared extension for php Any help would be greatly appreciated. Mike Sacauskis -- Much of the propaganda that passes for news in our own society is given to immobilising and pacifying people and diverting them from the idea that they can confront power. -- John Pilger ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: php_XML_ParserCreate error
David Landgren wrote: Mike Sacauskis wrote: Hi I'm trying to install mambo on a 5.4 release. I'm running into a problem with the mambo editor. I've traced it to an error in the apache log: /libexec/ld-elf.so.1: /usr/local/lib/php/20020429/xml.so: Undefined symbol php_XML_ParserCreate Hello Mike, I have just run into this same problem. Did you ever get the problem resolved to your satisfaction? I've just rebuilt php using a more recent version (4.4.0), and I'm getting the same error, and I can't see what's missing either. That should read 4.4.2 (upgrading from 4.4.0) David -- Much of the propaganda that passes for news in our own society is given to immobilising and pacifying people and diverting them from the idea that they can confront power. -- John Pilger ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Adding an extra Apache DSO module after 'make install'
List, Having built an Apache 2.0.58 from ports and watching it run, I realise that I forgot to include mod_negotiation. Is it possible to return the the ports directory, rebuild the package, this time with the addition module, and simply take the built mod_negotation.so and add it to the libexec/apache2 directory? Either that, or is there a more general method within the ports framework of taking a mod_*.c file and converting it to a DSO for inclusion with httpd after the fact? I.e, if possible I would like to avoid reinstalling the httpd world, but on the other hand, I don't want the new .so file to cause a segfault because of some sort of API mismatch. Thanks, David Landgren -- Much of the propaganda that passes for news in our own society is given to immobilising and pacifying people and diverting them from the idea that they can confront power -- John Pilger ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Apache not building its own modules (e.g. mod_alias)
Hello list, I wasn't exactly thinking things out, and just posted the following message to freebsd-ports, whereas it's probably better here. More eyeballs in any case. So sorry if you're seeing this for the second time. I'm having great difficulty tracking what's going wrong here. I'm trying to build Apache 2.0 with the following: cd /usr/ports/www/apache20 make PREFIX=/home/apache20 \ APR_UTIL_WITH_BERKELEY_DB=yes \ APR_UTIL_WITH_LDAP=yes \ WITH_MODULES=access alias auth auth_ldap headers info ldap mime proxy proxy_connect proxy_http rewrite status ssl \ WITH_STATIC_MODULES=access alias auth auth_ldap headers info ldap mime proxy proxy_connect proxy_http rewrite status ssl \ At the end of this, I get # sbin/httpd -l Compiled in modules: core.c mod_auth.c prefork.c http_core.c mod_mime.c mod_status.c mod_info.c mod_rewrite.c mod_so.c And no mod_*.so files created either. I can build some modules dynamically. For instance mod_access can, and I get a libexec/apache2/mod_access.so at the end. But some things just don't get built no matter what I try, such as mod_alias. The lack if mod_alias, for instance, causes Redirect to go unrecognized in the config file, and so on. I must be doing something wrong; any clues gratefully received. update: what I do see, grovelling around in the work/ directory is that mod_alias and a number of other modules appear to be explicitly disabled in configure.log and never enabled, although make show-modules show that it has been specified as enabled. For instance: (apache.port is just a wrapper around the above cd and make) ~/apache.port show-modules | grep enabled | sort -u access: enabled (static) alias: enabled (static) auth: enabled (static) auth_ldap: enabled (static) headers: enabled (static) info: enabled (static) ldap: enabled (static) mime: enabled (static) proxy: enabled (static) proxy_connect: enabled (static) proxy_http: enabled (static) rewrite: enabled (static) ssl: enabled (static) status: enabled (static) Thanks, David -- Much of the propaganda that passes for news in our own society is given to immobilising and pacifying people and diverting them from the idea that they can confront power -- John Pilger ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Apache not building its own modules (e.g. mod_alias)
David Landgren wrote: Hello list, I wasn't exactly thinking things out, and just posted the following message to freebsd-ports, whereas it's probably better here. More eyeballs in any case. So sorry if you're seeing this for the second time. I'm having great difficulty tracking what's going wrong here. I'm trying to build Apache 2.0 with the following: cd /usr/ports/www/apache20 make PREFIX=/home/apache20 \ APR_UTIL_WITH_BERKELEY_DB=yes \ APR_UTIL_WITH_LDAP=yes \ WITH_MODULES=access alias auth auth_ldap headers info ldap mime proxy proxy_connect proxy_http rewrite status ssl \ WITH_STATIC_MODULES=access alias auth auth_ldap headers info ldap mime proxy proxy_connect proxy_http rewrite status ssl \ arg, scratch all that. I got stung by not realising that /etc/make.conf takes precedence over what appears on the command line. Which seems backwards, but there you are. David -- Much of the propaganda that passes for news in our own society is given to immobilising and pacifying people and diverting them from the idea that they can confront power -- John Pilger ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: After upgrading Perl, cannot reinstall modules - Cwd.so: Undefined symbol perl_get_sv
Mark Kane wrote: Hi everyone. I finally did a big portupgrade on a 4.9-RELEASE system today and I'm having some problems with Perl. I got it updated to 5.8.7 (was 5.6.x before), ran use.perl port, and ran the perl-after-upgrade script which were all said in UPDATING. Now it's time to force upgrade/recompile the Perl modules for the new version. Unfortunately this is where I get stuck: xeon1# portupgrade -f p5-\* [snip] === p5-HTML-Tagset-3.10 depends on file: /usr/local/bin/perl5.8.7 - found === p5-HTML-Tagset-3.10 depends on file: /usr/local/bin/perl5.8.7 - found === Configuring for p5-HTML-Tagset-3.10 /usr/libexec/ld-elf.so.1: /usr/local/lib/perl5/5.8.7/mach/auto/Cwd/Cwd.so: Undefined symbol perl_get_sv *** Error code 1 On the odd chance that you haven't resolved this problem... Cwd is part of the PathTools perl package. Try forcing the reinstallation of ports/devel/p5-PathTools first and see if that gets things back in sync. David -- It's overkill of course, but you can never have too much overkill. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Fuser equivalent
Crucis wrote: Hi, Is there a BSD equivalent of the Linux fuser command? fuser is used to identify processes using files/sockets. Thanks in advance lsof? Available in ports David ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Stupid ASCII loader prompt
Svein Halvor Halvorsen wrote: [...] It's all about what connotation you put on the word. E.g. to make a file world read-writable you would type chmod 666 file. Ah, but you are talking in octal! % perl -le 'print oct 666' 438 No evil here :) Even though the number 666 is the number of the devil, the number itself is not evil. Just as little as the command is evil, or someone who types it. It's just a number. Put whatever meaning into it you like! Q.E.D. (my apologies for keeping such a silly thread alive) ObTopic: I think the OP was quite within his/her rights about wanting to disable the ASCII-art da?emon, yet wishing to continue to use FreeBSD. I find it pretty silly myself. I also note that FreeBSD understands 'tail -100 -F' and Linux (gnu fileutils?) doesn't. The devil is in the details, as it were. A chorus of reformat your disk and use something else you lowlife moron without a sense of humour does little to advance the cause. David ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Default security: other users can ACCESS MY HOMEDIR?!
Fafa Diliha Romanova wrote: hey i didn't realize all my users had full access to my homedir! that kinda sucks, me who thought i had everything private and locked down what chmod should i set my homedir to then? chmod 700 $HOME and how do i set my system to chmod all new homedirs to that chmod? umask 0077 thanks! You're welcome. David ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Different OS's? Marketshare
Anthony Atkielski wrote: Mike Hauber writes: [...] Well... There's a lot of options available. Personally, I prefer something like blackbox for administrative logins. It's _very_ lightweight and (like all things should be), you pretty much build it from the ground up. What do you mean by building it from the ground up? What do I get when I type startx by default? It looks extremely simple, whatever it is, just a few simple windows in green borders on a rather irritating gray crosshatched background. Yeah, ugly as f... fleas on a dog, isn't it? I found fluxbox to be a very nice window manager. Very lightweight, useable, and stylish enough to make people's heads turn and ask you what you're running. And the coolest bit is... tabbed x-terms! Just like Mozilla and web pages. No doubt other managers have this feature, but once you've used it it's very hard to go back. Fluxbox installs straight from ports, no trouble at all. I seem to recall there being two versions, ancient and current. Make sure you get the current one. David ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Anyone running Trend Micro IWSS in linux-compat mode?
David Landgren wrote: Hello List, I'm running a squid proxy on FreeBSD. The big problem today is web-borne viruses, spyware and other crap. The general feeling on the Squid users mailing list is that Trend Micro's InterScan Web Security Suite product is the way to go. Having looked at the various incomplete, non-functional free offerings, I have no problem going with a commercial product. The trouble is, they only offer it on Windows, Linux and Solaris. Just got the word back from their beta projects manager: Unfortunately, at this time, we do not have any concrete plans to support FreeBSD. Please feel free to visit the trendbeta website on a regular basis, as we often have new products (sometimes products are released based on consumer needs) listed there. Oh well, keep up the advocacy... David ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Anyone running Trend Micro IWSS in linux-compat mode?
Hello List, I'm running a squid proxy on FreeBSD. The big problem today is web-borne viruses, spyware and other crap. The general feeling on the Squid users mailing list is that Trend Micro's InterScan Web Security Suite product is the way to go. Having looked at the various incomplete, non-functional free offerings, I have no problem going with a commercial product. The trouble is, they only offer it on Windows, Linux and Solaris. So, is anyone running IWSS in Linux-compatibility mode? If so is there anything tricky to be aware of? I guess either way I'll try it out, and if the worst comes to the worst, I'll just have to reformat the beast and install Linux. But that would be a bummer. Thanks for any clues I can use, David ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: samba as wins-server
Florian Hengstberger wrote: Hi! I'm working in an office with several win hosts of all flavours (98,2000,eXPerience). Unfortunatly the resolution of computers takes sometimes up to half an hour (approx.) until they are accessible after booting up. In near future I'll have the chance to switch to FreeBSD with my box (at least, I hope so). I'll install samba for win access to my machine. Reading some documentation I've found out that samba can also act as a wins-server. Will this enhance the latency of netbios resolution or will it corrupt it? Are you the admin of this network or just a user? Is the network switched or hubbed? It probably won't get worse, but half an hour seems pretty excessive, there is something wrong. Is there a way to speed up this process with samba, am I writing complete nonsense? You will have to configure all the clients to direct their queries to the WINS server. If you're delivering addresses via DHCP, this can be communicated during address allocation, so that's not a problem. More of a hassle to update fixed-address machines, but even then modern Windows boxes no longer need to be rebooted for the changes to take effect. If this is just your machine and you don't admin the others it will have zero effect. Tell me if this is true. You might have other problems, I don't know. Have you run traces on your network to see what the traffic is? David ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: samba as wins-server
Florian Hengstberger wrote: I'm not the admin, I own a normal workstation. I just wanted to know if setting up samba as an wins-server improves the speed of netbios resolution. It seems to me that I'll get in trouble when other XP boxes are around, because I've heard that the first win XP box on the net rules the workgroup an manages all netbios stuff. By the way: If samba conflicts with XP, how do XP machines manage not to get in trouble if there is more than one XP box? Your admins probably already have a WINS server, you should point to it and you'll be set. Look at the 'wins server' parameter in smb.conf. As to how XP boxes get on with each other, they hold elections, and the host with the most votes wins (more or less). The exact parameters that control the outcome are: os level preferred master domain master local master Samba documentation explains this pretty well. David ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Different OS's? Marketshare
Robert Kim, Wireless Internet / Wifi Hotspot Advisor wrote: Different OS's? Marketshare... any idea how many major OS's are out there today and what market share they have? i think WIN 70% Lin 20% Apple 5% Where did you get these numbers? so who is the other 5 % ??? Well, other than *BSD, names like Solaris, HP-UX, AIX and Tru64 spring to mind. David ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: perl and ports
Kris Kennaway wrote: I will start with the cvsup developer :P cvsup itself doesn't require perl to build (as you can see from the lack of mention in the makefile), it's one of the other build dependencies. Well I think Modula-3 is find totally useless and even obsolete and dead as far as languages go. Except that it is needed to build cvsup :) David ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Creating your own FreeBSD install CD-ROM
List, how does one go about creating a FreeBSD install CD-ROM? I'd like to be able to build my own CD-ROM with as much as possible of the install process done automatically. For instance, what parts of FreeBSD to install: X or not, games or not, etc. etc. And which ports to install straight away. I could also set up cfengine so that as soon as the machine came to life it could go and connect to my cfengine repository and download all the latest configurations to move into operational status. E.g., I'd like to just have to specify the partition layout, the name and IP of the machine, and its role(s): (web, dns, squid, mta, samba...) This would make rebuilding so much faster if a machine was catastrophically trashed. And it would make my computing installation much less dependent on me. Thanks for any pointers to documentation on the subject, I'm sure people have done this before, but I can't coax the results out from my web searches. Thanks, David ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Large file on a DVD-ROM showing -ve size
Folks, I know FreeBSD deals with files larger than 2Gb, but I have a large file on a DVD I can't copy. This is on 4.10-STABLE FreeBSD, compiled Tue Oct 5 09:42:59 CEST 2004 ls -l gives: -r-xr-xr-x 1 root wheel -2131373162 Nov 17 01:05 bas1.bas The fact that ls itself gets it wrong makes me nervous. I compiled a 5.8.6 Perl with large file support and 64bitint and ran the following: /usr/local/p586-64i/bin/perl -le \ '$_=shift; print $_\t, (stat $_)[7]' bas1.bas bas1.bas-2131373162 I updated my copy of rsync from ports and tried that way: % rsync -av /cdrom/bas1.bas /home/david/ building file list ... done bas1.bas sent 87 bytes received 40 bytes 254.00 bytes/sec total size is -2131373162 speedup is -16782465.84 ... and the resulting file is 0 bytes. I'm wondering if there is something special I have to add as a parameter to the mount command, in order to have the kernel interpret things differently. I've had a look at mount_cd9660 but nothing leaps out at me. The mount currently looks like this: /dev/acd0c on /cdrom (cd9660, local, read-only) Or something else? Thanks for any clues I can use, David ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: SSH 5.3 Problems
Brian McCann wrote: Hi all. On 5.3 and 5.3 RC1 I have this problem where when I ssh in using either a FreeBSD 4.3 box or an older PuTTY client (0.52 is one I experienced it with), I cannot connect. On PuTTY, it asks for a username, then just exits. On FreeBSD when I put the ssh client into verbose, it appears it can't agree on an auth method, even though both are set to accept and attempt keyboard-interactive. Upgrading to a newer PuTTY fixed the problem, and it seams to work from FreeBSD 4.10. I've seen some posts from people that upgrading their PuTTY fixed the problem, but I'm curious if anyone knows the cause of the problem, and possibly how to fix it, on the server end. What error message does sshd emit on the server? David ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: bash - superuser
Giuliano Cardozo Medalha wrote: Hi, I have a machine with FreeBSD 5.3 - release -p2. I have installed bash from ports. How is possible to use bash in root account ? Thanks a lot Don't. Leave /bin/sh as your shell. If you want to run bash as root, log in as usual and then run 'exec bash' to replace your current shell with bash. This is a basic rule of hygiene when working as root. You'll appreciate working this way when you lose your /usr partition one day... David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Change Bash-3.00# Prompt???
Adam wrote: I installed bash shell and now my prompt says bash-3.00# How do I change the text before the #? I'd like it to say bash# instead of bash-3.00#. Edit .bashrc or .bash_profile. Run 'man bash' and look at the FILES section. David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: installing bsd on a laptop
[EMAIL PROTECTED] wrote: hello, do you know of any laptop brands that can run freebsd or openbsd that is available to purchase??? Available to purchase, I dunno. With older HP laptops I had no success with 5.2, but 4.10 was fine. I haven't tried 5.3 yet. When I say no success, I meant that the laptop refused to complete the boot of an installation CD. If you can get to a shop, take an install CD along with you. If you can boot up and get as far as the main menu, it'll probably be just fine. David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: bash - superuser
Dick Davies wrote: * Gerhard Meier [EMAIL PROTECTED] [1207 12:07]: On Mon, Dec 20, 2004 at 08:41:57AM -0200, Giuliano Cardozo Medalha wrote: I have a machine with FreeBSD 5.3 - release -p2. I have installed bash from ports. How is possible to use bash in root account ? Do not change the shell of the root account. If you have /usr or /usr/local on a separate partition, and you cannot mount for some reason, you wont be able to fix that, without booting from another device. No, but you'll still be able to use /bin/sh when going single user, so what's the big deal? I really don't get what the problem is with this 'sh is on the root' argument. Using bash is a lot more productive for many people, so why not let them use it? If you're really terrified of not knowing how to use sh, then stick a static bash in /bin. To the original poster: just be root and run 'chsh'. No. When you are logged in as root, you *should* have to go through extra hoops to get comfortable. I am not saying that you should not use bash when logged in as root. I am saying that you should not configure your root account to login with shell that is dysfunctional if /usr is unmounted. Yes, 'exec zsh' or whatever is a minor hassle, but it's there to remind you that root is different. If the OP had to ask, then it's pretty clear that he shouldn't. David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: bash - superuser
Joshua Lokken wrote: On Mon, 20 Dec 2004 12:29:37 +0100, David Landgren [EMAIL PROTECTED] wrote: [...] Leave /bin/sh as your shell. 'Leave' /bin/sh as your shell makes it sound like /bin/sh is the default root shell. Did this change in FreeBSD 5.x? It appears that in 4.x, the root shell is /bin/csh by default, which [I believe] is linked to /bin/tcsh. No, it's still /bin/csh. My bad. I hate csh so much I usually change it to /bin/sh :) David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: CVSUP Routing question
Mark Jayson Alvarez wrote: Good day! Is it possible to tell cvsup to use another machine's global access in fetching the freebsd source updates?? Here's my office workstation setup: (private ip) (pri/pub ip) (all public) workstation router proxy server---internet mail server web server [...] I need to update our private LAN workstations using CVSUP but I don't know how exactly will I do it. Any idea? I run run my own cvsup ports mirror on a perimeter box, what would be your public web server. Hint: look into cvsupd All my internal machines cvsup off the perimeter machine, so the upstream cvsup provider is hit only once by me. IOW, their cvsup.ports files make reference to my box, not cvsup.foo.freebsd.org David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: squid-downloads
metallarch wrote: -- How can i deny downloads from squid? Here's a novel idea, how about reading the documentation? http://www.squid-cache.org/Doc/FAQ/FAQ-10.html ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Finding options for ports
[cc list trimmed] Lowell Gilbert wrote: Paul Schmehl [EMAIL PROTECTED] writes: --On Monday, November 15, 2004 06:46:42 PM -0600 Conrad J. Sabatier [EMAIL PROTECTED] wrote: This is a useless use of cat. You could accomplish the same thing with: grep WITH Makefile When there are ten different ways to skin a cat, what makes one way inherently better than another? Efficiency. Using cat(1) spins off a subprocess. http://sial.org/howto/shell/useless-cat/ ooh, just saw this message while catching up on the list. rant This is a pet peeve of mine. Yeah, sure it spins off another sub-process, but who's counting? A more important issue is that of comfort. Your comfort. If you have a number of strings that you want to grep for, it's much more cumbersome to work back past the files at the end of the command line to change the grep target string. grep foo somefile grep bar somefile Assuming bask, you have to step over 'somefile' with some whacky Esc-B combo in order to ^W zap the 'foo' field. Contrast this with: cat somefile | grep foo cat somefile | grep bar In this second case, you can recall the command and ^W to zap the token immediately, and replace it with bar. Much less typing effort. The different strings you grep with, the more you save. And then once you've grepped what you want, you might then replace the grep by sed, awk, wc or whatever. (Yes, I know about ^old^new, but that is even more cumbersome that Esc-B). Let the damned computer burn another process. That's what it's good at! /rant David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: can't get rid of this file with trailing backslash?
Bart Silverstrim wrote: [...] I'm too paranoid that I know what *should* work wouldn't or would still end up deleting the original file I wanted, so I'd have to make a backup of the file and do it that way rather than play with escapes and quotes. Cant' you escape the \ with a \? rm named.conf\\ ?? I think he did do that and it worked. I was just commenting what my first instinct is to do. A few extra keystrokes, but it saves my peace of mind. I jump among too many o/~ speaking words of wisdom o/~ I couldn't agree more. Another thing that no-one else has yet mentioned is rm -i named.conf\\ which will force rm to prompt for a y/n response in order to proceed with the actual unlinking. David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Sluggishness on a 6 CPU HP-6000 Netserver
Hello,
I have just installed an HP-6000r Netserver with 2Gb RAM and 6 P-III
700MHz processors.
I cvsupped src (4.10-STABLE) yesterday and built the world and kernel.
I've stripped out just about everything else I can from the kernel: usb,
firewire, apm, serial port, //l port, PCMCIA, other RAID/SCSI/network
drivers.
The kernel has dummynet support. I have tried option HZ=x, for x=1000,
200, 100 but network response is sluggish. I haven't specifically
referred to any dummynet features in my ipfw ruleset.
When I type quickly I can run several characters ahead of the display,
which jerks along behind. Display output, e.g. when compiling the
kernel, is also jerky. From the console the response is crisp.
Both the switch to which the box is connected (a Cisco Catalyst) and the
fxp network driver are hard-coded to 100baseTX full-duplex.
netstat -m:
70/480/26624 mbufs in use (current/peak/max):
66 mbufs allocated to data
4 mbufs allocated to packet headers
64/128/6656 mbuf clusters in use (current/peak/max)
376 Kbytes allocated to network (1% of mb_map in use)
0 requests for memory denied
0 requests for memory delayed
0 calls to protocol drain routines
It must be something, but I have no idea what to look {at,for}/what to
tune. It's the first time I've built dummynet, first time I've used a
box with more than two CPUs and I wonder if there might be some sort of
contention happening.
Thanks for any clues I can use.
David
___
[EMAIL PROTECTED] mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-questions
To unsubscribe, send any mail to [EMAIL PROTECTED]
Sluggishness on a 6 CPU HP-6000 Netserver
Hello,
I have just installed an HP-6000r Netserver with 2Gb RAM and 6 P-III
700MHz processors.
I cvsupped src (4.10-STABLE) yesterday and built the world and kernel.
I've stripped out just about everything else I can from the kernel: usb,
firewire, apm, serial port, //l port, PCMCIA, other RAID/SCSI/network
drivers.
The kernel has dummynet support. I have tried option HZ=x, for x=1000,
200, 100 but network response is sluggish. I haven't specifically
referred to any dummynet features in my ipfw ruleset.
When I type quickly I can run several characters ahead of the display,
which jerks along behind. Display output, e.g. when compiling the
kernel, is also jerky. From the console the response is crisp.
Both the switch to which the box is connected (a Cisco Catalyst) and the
fxp network driver are hard-coded to 100baseTX full-duplex.
netstat -m:
70/480/26624 mbufs in use (current/peak/max):
66 mbufs allocated to data
4 mbufs allocated to packet headers
64/128/6656 mbuf clusters in use (current/peak/max)
376 Kbytes allocated to network (1% of mb_map in use)
0 requests for memory denied
0 requests for memory delayed
0 calls to protocol drain routines
It must be something, but I have no idea what to look {at,for}/what to
tune. It's the first time I've built dummynet, first time I've used a
box with more than two CPUs and I wonder if there might be some sort of
contention happening.
Thanks for any clues I can use.
David
___
[EMAIL PROTECTED] mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-questions
To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: max group name length
David Bear wrote:
was just trying to determine the maximum string length of a group
name.
found
struct group {
char*gr_name; /* group name */
but no size.
any pointers (with limits)?
Hmm, I had a browse through the kernel source for a while but didn't
find anything definite. What I do know is that you should endeavour to
keep the length no greater than 8. Up to 15 is probably ok as well on
modern kernels, and group names longer than 15 is getting a bit silly.
Sorry no to have anything more precise.
David
___
[EMAIL PROTECTED] mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-questions
To unsubscribe, send any mail to [EMAIL PROTECTED]
'closed stream' on portupgrade, pkgdb -F
List, I'm having trouble bringing a machine up to par. I migrated from 4.8-RC2 to 4.10-STABLE this morning. No problems there. When I do a pkg_version -vL= I get a couple of dozen things to upgrade. I've synched my ports tree, run pkgdb -Uu. I have rebuilt ruby18 % ruby --version ruby 1.8.1 (2004-05-02) [i386-freebsd4] and portupgrade and yet each time I run the following commands % pkgdb -F --- Checking the package registry database closed stream % portupgrade png closed stream If I make deinstall make reinstall portupgrade I get % portupgrade png [Updating the pkgdb format:bdb1_btree in /var/db/pkg ... - 76 packages found (-1 +1) (...). done] closed stream I guess I can run around manually and update stuff, but it is a bit of pain. If anyone has a clue, I'd be most grateful. I don't see what I'm overlooking. thanks, David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Is there a Complete Package(NOT Ports) for: [Apache+PHP+MySQL+Mod_SSL+Mod_Perl] ?? - newbie+3
Henrik W Lund wrote: DK wrote: Hi all, I just wanted to know if there is available anywhere a Complete Package that is Ready to Go for a FreeBSD Server imlementation that contains: Apache PHP MySQL Mod_SSL Mod_Perl ... before anyone flames, yes I know you can just d/l the ports compile from source! ... Thats not what I am looking for. Tried that found it was too much messing around (compared to the equivalent on Windows 2000) Um, there is no equivalent on W2K, but I digress... One thing to remember is that Apache and mod_perl are usually run as two separate instances. You run static Apache off port 80, and dynamic Apache (hence +mod_perl) on port 8000 or whatever, and this is reverse-proxied back through the static server at port 80 (and 8000 is firewalled off from the outside). This results in big savings in memory. mod_perl processes can weigh 80Mb apiece, easily. Standard httpd processes weigh in at about 3Mb. You don't want to tie up an 80Mb process spoonfeeding some luser on a modem. Reverse proxying lets you have a large number of small static processes and only a handful of mod_perl processes. There are also a number of ways you can proxify the dynamic requests, so there's no one-size-fits-all. The easiest way is to do something like: ProxyPass/dyn http://localhost:8000/ ProxyPassReverse /dyn http://localhost:8000/ Then you have to decide who gets the https port. My advice would be let the static handle it. (--enable-ssl) Next, the question of MySQL. I expect you mean that PHP and/or mod_perl needs MySQL support. PHP configure will detect this automatically if you have it installed. In mod_perl, you'll probably want to attack it via Apache::DBI and DBD::Mysql. Finally, you can install mod_perl and PHP and DSO (i.e., loadable modules), after you have your basic installation up and running. (Just remember to --enable-so). The level of PHP processing you plan to do will indicate whether you should be running it directly out of port 80 or proxified from 8000. There are so many ways to put these pieces together, depending on your needs, that there is no simple package that puts them all together. I sympathise with your remark that it's so much easier under W2K. The trouble is, a well-tuned mod_perl installation will run rings around anything IIS/ASP can do, but that comes at a cost of knowing how to set it up. I would recommend investing in the mod_perl book by Stas Beckman and Eric Cholet. http://www.modperlbook.com/ Regards, David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Increase Inodes
Rus Foster wrote: No. Changing the number of inodes means recreating the filesystem, and that implies wiping out the contents. I thought as much. Time to start doing a backup.. Either that or see if there isn't a pile of cruft hidden away in some directory that you forgot about. Such a thing happened to me once and by deleting a pile of stuff I was able to limp along quite well until I was able to schedule a rebuild. David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: mail daemon(s) recomendation request
Mike Maltese wrote: I'm looking for an smtp daemon that is less complex than sendmail, can use procmail and has some virtusertable equivalent features. Allowing relay based on mx is going to be a requirement for some of the boxes as well. Sendmail is working ok for me, but I'm intersted in finding something with a config file I can actually read. I'm going to check into postfix and maybe qmail. I'm also looking for some imap and pop3 daemons (ideally with ssl support) that can use maildir. I'll probably continue to use dovecot for imap and pop3. Anyone have any other suggestions? Qmail is said to be a very robust and secure MTA. Exim is also a popular Sendmail replacement. Personally, I swear by Postfix. Stable, secure and easy to configure, with lots of back-end options. For pop3/imap, I like Courier-IMAP. I'll second the Postfix recommendation. It can do maildir style mailboxes if you want, does procmail, and has good virtual user support. Not sure what you mean by relay based on mx, but Postfix's configuration allows for all sorts of weird setups if needed. David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED] -- Commercial OS breeds commerce, whereas free OS breeds freedom, the only thing more dangerous and confusing than commerce. -- Michael R. Jinks, redhat-list, circa 1997 ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
failure installing 5.2.1: cbb0: Unsupported card type detected
Hello, I have an old laptop, an HP XE3 OmniBook. Nice machine from an ergonomical point of view, and I was thinking of recycling it by installing FreeBSD. I downloaded the ISO for 5.2.1. It starts to boot just fine, but during the kernel load it dies with the message cbb0: Unsupported card type detected I tried selecting Safe mode, but it freezes earlier with cbb0: [MPSAFE] A few lines above it says cbb0: 02Micro 026933 PCI-CardBus Bridge irq 9 at device 4.0 on pci2 Is there any way to get around this? I've seen a few messages on the subject in the archives but apparently these all involve running systems, something I don't yet have :( Or should I try 4.10 instead? Thanks for the clues, David -- Commercial OS breeds commerce, whereas free OS breeds freedom, the only thing more dangerous and confusing than commerce. -- Michael R. Jinks, redhat-list, circa 1997 ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
ps: warning: /var/run/dev.db: No such file or directory
I recently rebooted a server that had been running for many months. I haven't touched the kernel or userland programs since it went into production. The server was rebooted with 'shutdown -h now', powered down, and then later restarted. I've since noticed that cron didn't restart, which is odd, but fixable, but more importantly, when I run ps, it spits out 'ps: warning: /var/run/dev.db: No such file or directory' (although, as far as I can tell, the output is perfectly reasonable). I'm wondering if one is a symptom of the other. In any event, /var/run/dev.db is most certainly not there. I guess I could reboot the server tonight, but I'm not sure that that will fix it, as I don't understand the cause. I've searched the archives a bit, and the best thread I could find dated from 1997, and suggested that it could be due to an unclean shutdown, which is definitely not the case here. I'm running FreeBSD 4.8-STABLE, where stable is defined as being what it was around June 2003. I'd be grateful for any pointers you might have. Thanks, David -- Commercial OS breeds commerce, whereas free OS breeds freedom, the only thing more dangerous and confusing than commerce. -- Michael R. Jinks, redhat-list, circa 1997 ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Problems with ssh
Peter Ulrich Kruppa wrote: Hi! I can't ssh into one of my machines anymore. The only possible reason I can think of is that I played around with it's IP. Does it have to be reset somehow? Delete any references to it in ~/.ssh/known_hosts on any machine from which you wish to connect to the machine in question. Also, try ssh -v ... to see why/when it fails. David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: exe order in /usr/local/etc/rc.d
Konrad Heuer wrote: On Thu, 16 Oct 2003, RJ45 wrote: I have never understood in which order are execurted the scripts which are in /usr/local/etc/rc.d most port applications like mysqld, cyrus-imapd, spamd etc. put their startup scripts into /usr/local/etc/rc.d but they are not ordered by a number attached before the script name (like the Unix SysV style) anyone has some hint about this ? How can I Set the startup order into /usr/local/etc/rc.d [...] Thus the scripts are executed in alphabetic order. Which means you are free to rename them as 00-mysqld.sh 20-cyrus.sh ... etc, in order to obtain the necessary ordering. Just remember that if the files were put there from ports, upgrading or removing the port in question, expect it to squawk about missing files. If you were really fussed about it, you could create a /usr/local/etc/rc.d/ordered, and symlink /usr/local/etc/rc.d/mysqld.sh to /usr/local/etc/rc.d/ordered/00-mysqld etc. etc. as required and set local_startup=/usr/local/etc/rc.d/ordered in /etc/rc.conf David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Increasing per-process memory limits?
Giorgos Keramidas wrote: On 2003-10-14 23:51, njc [EMAIL PROTECTED] wrote: Hello, I'm currently using Zope's Plone application which relies heavily on python. When uploading large files (65 megs, possibly higher), the python process is bombing out with a 'memory exceeded' error. I'm using 4.8-STABLE . I've been told that if I increase the amount of available memory to the python process that I can resolve this issue. I've attempted increasing the max users variable in the kernel to 0, hoping it might auto-tune, but that doesn't seem to be working. Am I tweaking the wrong area? Where should I be looking? Try reading the manual page of /etc/login.conf: man login.conf Another way of enforcing/changing the limits that a user process has is through the 'limit' built-in command of tcsh or the /usr/bin/limits system tool. More information about these in the tcsh(1) and limits(1) manpages. Note that the GENERIC kernel has a hard limit of 256MB, defined by MAXDSIZ (and MAXSSIZ too) in the configuration file. For modern machines with a large amount of RAM this may be inadequate. You will have to recompile the kernel to run larger processes. David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: FreeBSD SPAM
Robert Huff wrote: Roger 'Rocky' Vetterberg writes: SpamAssissin can hog a lot of CPU if you handle a lot of emails, so make sure you are running it in daemon mode, that helps quite a bit. SpamAssassin is Perl, so of ourse it's a hog. I seem to remember someone trying to write a version in C Well, they'll still be trying in 2007... BNy the time you've written the current SA in C, the current Perl version will be miles ahead. Programmer productivity is always more important than raw power. A 2.8MHz Pentium Xeon with 2Gb RAM doesn't cost all that much and offers phenomenal processing power. If that's not an option, an effective method for reducing SA load is to feed it less email :) I use Postfix and have some pretty extensive correlation checks to filter out spam (spoofed sender domains, garbage HELO strings, obsolete or spambait recipients, spammer hosts). Since the beginning of the month these low-overhad checks blocked 5313 messages of 19990 total. That's 5300 less that have to be dealt with SA. To learn more about blocking spam with Postfix, http://www.securitysage.com/guides/postfix_uce.html is a good place to start these days. David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: cvsup - down?
Bill Moran wrote: Hey all. This seems like a really dumb question to me, but I'm having such a rough last two weeks that I'd rather be sure by sounding stupid, than by looking stupid. What I'm doing is upgrading a machine to RELEASE_5_1 (to take care of the recent arp problems) and I wasn't paying attention when I started cvsupping, and accidentally started it cvsupping to HEAD instead. I realized my mistake halfway through and canceled it. Then restarted it with the proper tag=RELENG_5_1 I'd throw it all away and start again as you suggest as a second alternative. This assumes you have a set up like I have: one machine that cvsup's against an official cvsup host, and then that machine is used the source for the rest of the machines. Makes the cvsups go really fast. Later, David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Deleting/Filtering mail from POP3 server before its download?
ekrem _ wrote: [...] I'd like to be able to delete my emails from my ISP's POP3 server, especially to avoid the recent bombardment of the W32.swen.A virus which come with a 150kb+ attachment. So I'm looking for such a filter or spam detector which will delete from spam directly from the POP server. I can't vouch as to whether it works because I don't use POP, but the following program was posted to Perlmonks yesterday. http://www.perlmonks.org/index.pl?node_id=292982 David Thanks David, and to Bart Lateur (the author of the perl script), I'm using the perl script from the above link and it's working fine. It's already saving me download time and bandwidth. Great news. You should post to the page in question and relate your success. You don't have to create an account, Perlmonks supports slashdot anonmymous-coward-type posts. I'm sure other people would like to hear confirmation that the scripts works from a third party. David :x ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Can I disable HTT in 4.8-STABLE?
Kent Ketell wrote: Is there a way to disable the hyperthreading in 4.8-STABLE or 4.9-PRERELEASE? I'd like to be able to test with and without HTT enabled to see whether it's actually helping or not. You have to compile a kernel with and without the options SMP, APIC_IO and HTT. You can test if you're running with HTT by examining the output of 'sysctl hw.ncpu'. David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Emulating gmake functionality in make
Hello list, gnu's make (a.k.a gmake) has a nifty piece of functionality that, while it may be possible to perform with make, for the life of me I can't figure out the syntax. Let's say you have a list of files a b c d that get transformed by program foo into a.x b.x c.x and d.x Further, let us suppose that you have another list of files e f g that get tranformed by bar into e.x f.x and g.x Note how the resulting files all have a .x extension, yet are produced with two different programs. In gnu parlance, one would write something like FOO = a.x b.x c.x d.x BAR = e.x f.x g.x $(FOO): %.x: % foo $ $(BAR): %.x: % bar $ Which roughly translates to for the files in the FOO list, you get to something.x from something, by running 'foo something' (since foo produces something.x as a result). And the resulting Makefile is very compact. Can this be done with make (without repeating things in more than one place in the Makefile)? I'm quite happy with the gmake way of doing things, except of course I keep running 'make' instead of 'gmake' and it complains out the syntax. If I could just teach make how to do this it would be more comfortable. Thanks, David :x ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Deleting/Filtering mail from POP3 server before its download?
ekrem _ wrote: Hi people, I'd like to be able to delete my emails from my ISP's POP3 server, especially to avoid the recent bombardment of the W32.swen.A virus which come with a 150kb+ attachment. So I'm looking for such a filter or spam detector which will delete from spam directly from the POP server. I can't vouch as to whether it works because I don't use POP, but the following program was posted to Perlmonks yesterday. http://www.perlmonks.org/index.pl?node_id=292982 David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Hmm! I couldn't move the old kernel over! during upgrade
Hello, I'm trying to recover a machine that failed to upgrade from 4.8 to 5.1 (the make installworld failed with a sh core dump half way through :( The restore program is hosed, so I can't restore from tape. I don't have a recent backup of /home either (it's a hardware RAID-1, so I'm slacker than I should be about backups...) so I don't want to do a fresh install. I tried to do an upgrade from the original CD I used to install it. Things go pretty smoothly, I'm able to identify the fs partitions, and select what I want to install, but after the partitions are fscked, the install spits out the message Hmm! I couldn't move the old kernel over! And asks if I want to reboot. On the Alt-F4 shell, I can see my partitions mounted under /mnt, and I can see the kernel and modules. I can move them around (and renamed them: kernel - kernel.x) but still I get the same error message. Has anyone had this sort of trouble before, and if so, what's the fix? Thanks, David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: OT: finding a callers number
Jonas wrote: Sorry for the OT but I have an urgent question and don't really know how else to ask. We have some hostile issues in our planned community against our kids and need to find out who called us between 5-8PM Sunday evening. My 3 1/2 year old picked up the phone call and I need to find out who the Your telephone is running FreeBSD? Cool! caller was. Our local phone company is SBC. Unfortunately I don't have caller ID yet (will tomorrow) so I need to get the callers number from the phone company. Since I can't subpoena SBC for the records is there any other way to get the callers number? Maybe in /var/log/phone? ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: FW: Spam mail warning notification! (MS-Update)
Schalk Erasmus wrote: It is interesting to see that some of our colleagues belonging to the FreeBSD Mailing List, are infected with the Worm.Swen virus! Sven forges envelope senders; that you receive messages purporting to come from FreeBSD lists means nothing. I'm leaving this list, since I get hundreds of these DAMN MICROSOFT update emails. Luckily you can filter these stuff using TREND InterScan. Whatever works for you... David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: postfix, MailScanner, and SpamAssassin on fbsd 5.1
dave wrote: Hello, I've read the installation howto on the MailScanner home page and have got postfix and MailScanner to talk to each other. The problem is whenever i enable the use spamassassin option and restart MailScanner i start getting errors in my maillog about messages being found but no queue directories and to enable hash_queue_names and hash_depth in postfix. I do this and it does not stop the problem, i eventually would also like to add a virus checker, which one is recommended for use on fbsd 5.1? Any help appreciated. dO NOT use Mailscanner with Postfix. Wietse does not like external programs coming in and fiddling with Postfix's innards. Use Amavis if you want to add a virus scanner. David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: *how to obtain a better result from this list*
Paredes Sánchez Martín A. wrote: Hi: I am looking for a mail that says how to obtain a better result from this list. Does someone can send to me? I forwarded How to get best results from FreeBSD-questions, so there's no need to spam him with 10 copies :) David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Sendmail is sleepy
Charles Howse wrote: Did you enter your hostname and ip address in /etc/resolv.conf? I don't if it will fix your problem; but is helps when many applications complain about the hostname. I think you mean /etc/hosts, something like 10.0.0.1 larry.domain.tld larry Andrew Gould Uh, no I didn't. Man resolv.conf doesn't say anything about how to do that, and it's never been necessary. Putting the above in /etc/hosts is good insurance, if ever your DNS servers (the addresses of which are specified in /etc/resolv.conf as it happens) go out to lunch. You have to consider the trade offs between always knowing the IP addresses of remote machines when a packet absolutely, positively has to get there when your DNS is down, balanced against the likelihood of those machine changing IP addresses (upgrades, usually). Which means visiting all your servers and updating their hosts files. David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Huge processes (was: Re: Large memory issues)
Irvine Short wrote: On Wed, 27 Aug 2003, David Landgren wrote: Irvine Short wrote: I then found that this: options MAXDSIZ=(2048*1024*1024) options MAXSSIZ=(128*1024*1024) (and also 64MB) options DFLDSIZ=(512*1024*1024) worked fine but not as expected - limit reports datasize unlimited. I've managed to crank it up as far as options MAXDSIZ=(3568*1024*1024) Cool! Although I tried 3500 it blew up too... Bah! Just for the record, I read those figures from a discarded kernel configuration file, and the above values (3568MB max data size for a process) don't work. The highest value I've been able to boot correctly with is (only) 2816Mb. There's probably a few more megabytes that can be eked out, but I'm pretty sure that 3072 fails. The error is something to do with the kernel being able to map the largest-sized process into memory. It blows up at boot-time with some sort of vmalloc or kmalloc error. I searched the archives when I was working on this, and came across a kernel developer who replied to someone having simuilar difficulties and they said why would you want to do something like that? His reasoning being that the left-over memory would be better used by the OS for caching and buffering anyway. I don't really consider that as a good answer, at least not for modern machines with large amounts of RAM. I have a 4Gb RAM server running Squid, and nothing else. Squid represents a very specialised problem domain and has elaborate algorithms to decide what to keep in RAM, and what, and when, to write out to disk. Much more so than the OS, which is tuned to deal with the common case. When it's time for an object to be written out to disk, it should be written out quickly, so that the RAM can be freed up and given to something else more deserving of being cached. As it happens, the SCSI controller has a large slab of RAM on it too, so there's even less point for the OS to hold onto it for too long in disk buffers. As it is, I never see the Cache and Buf values in top(1) rise about 85M and 199M respectively. I take that to mean that the OS isn't using the extra memory either. I've looked around at sysctl settings and the source and the documentation suggests that modifying anything to do with VM settings is akin to meddling with the affairs of wizards. This seems to me then, when adding in another 20Mb for sundry housekeeping processes, that just under a gig of RAM is going wasted. I could easily cache another 5 web objects *in RAM* if I could make it available to Squid. So if there's something that can be done about huge maximum process sizes, I'd love to hear about it. I'd *really* like to be able to have a 3.5Gb process in a 4Gb machine. 512Mb ought to be enough for everything else. I've become skilled in not running anything else on that server that could possibly chew up RAM and upset Squid. David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: File permissions suddenly change for /dev/null.
Ed Alley wrote: I'm running FreeBSD-4.8. Sometimes the file permissions for /dev/null get mysteriously changed by some unknown process to: crw--- 1 root wheel 2, 2 Sep 2 11:20 /dev/null This has a devastating effect on user processes that want to open /dev/null. Whenever my system starts acting funny, the first place I look is at the permissions for /dev/null. When I find them changed I go under root and execute: chmod 666 /dev/null to get things back to normal. Has anybody seen this before? Have I got a hidden umask set up wrong somewhere, or is one of my daemons the culprit? Or could it be happening during the time that I run as root doing system maintenance? The only time I've had something like this happen to me was when I installed Frontpage Extension for Apache. Come to think of it, it was on a Solaris box, but no matter. Apache has two vestigal config files that no longer serve any real purpose, srm.conf and access.conf. There are two schools of thought about what to do about this. The first school says to have those files sitting in the filesystem, but empty. The second says that the configuration directives in httpd.conf that refer to those files should point to /dev/null instead, and not have any other config files. I subscribe to the latter idea: it makes for a cleaner conf/ directory. Unfortunately, Frontpage Extensions, in Microsoft's oh-so-typical couldn't give a fuck about anyone else way actually wants to put some kind of garbage in the access.conf file. Morons, the rest of the world hasn't been doing that for at least six years by my count. Anyway, rather than bombing out, the script managed to mangle /dev/null utterly, and turned it into -rw-rw-rw 1 root wheel 2, 2 Sep 2 11:20 /dev/null I.e., a *plain* file! This symptoms this produced were really bizarre. As it turned out, /dev/null on Solaris is a soft link to the real device in /device (which may be why the FPE install script was able to trash it). It was easily repaired by restoring the symlink. Anyway, the moral of the story is, what have you installed recently? David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Newer names
Erik Trulsson wrote: On Sat, Aug 30, 2003 at 04:13:20PM +0200, Hendrik Hasenbein wrote: Vitali Malicky wrote: What about DRAGON? There is also such a term as DRAGON. Yes, I'm all serious. cron is a DRAGON, for example. DRAGON is a process which may run other processes periodically under a user specified. What about it? And many my colleagues, and me as well, refer to it also as PROCESS. You've made me curious. Can you explain me the acronym DRAGON? Or did you just make that up? 'Dragon' is not an acronym. It is just a name. 'Daemon' is not an acronym either despite what some people tell you. The expansion 'Disk And Execution MONitor' was made up after the term 'daemon' had come in use, and should not be used. I believe the correct term is a backronym :) David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: How to run PostgreSQL on boot?
Alex Zivenko wrote: Hi all! I need to run Postgresql server on startup. It means, that I need to start postmaster every boot. How can I do this? Sorry for my english. I have the following file on the servers that run Postgresql, named /usr/local/etc/rc.d/postgresql.sh: #! /bin/sh su postgres -c '/home/pgsql/bin/pg_ctl start \ -D /home/pgsql/data -l /var/log/postgresql' You will probably have to change paths etc. to suit your site, but this is the general idea. David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Off Topic RegEx Question
Roger Williams wrote: I know thins is not the place but I know one of you know this one off the top of your head. I have: $list = dog 1 1 1 cat 2 1 snake 111 and I want to end up with: dog 1 cat 2 snake 1 I thought $list =~ s/ \d \d/ \d/g; would do the trick, but that gives me: dog d 1 1 cat d snake d 1 \d in the RHS of the s/// doesn't do much (as you can see...) Try: $list =~ s/(\d)\d*(?: \d+)*/$1/g; Capture a digit, maybe followed by more digits, then followed maybe by groups of space and digits. Replace all that by the captured digit. Note that this will transform: dog 1 4 7 cat 2 1 snake 123 = dog 1 cat 2 snake 1 rather than: dog 1 4 7 cat 2 1 snake 123 = dog 7 cat 2 snake 3 I assume since cat 2 1 = cat 2 that you always want the first digit matched. http://www.perlmonks.org/ is a good place to ask Perl questions. David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Setting sysctl variables BEFORE the kernel boots and runs init?
Micheal Patterson wrote: - Original Message - From: Donald Burr [EMAIL PROTECTED] To: FreeBSD Questions List [EMAIL PROTECTED] Sent: Wednesday, August 27, 2003 6:11 PM Subject: Setting sysctl variables BEFORE the kernel boots and runs init? I would like to be able to set certain sysctl variables in the kernel, preferably BEFORE the kernel finishes booting and runs init, etc. (or at the worst case, very shortly after init runs) I thought I remembered that there was a way to do this through the boot loader. Can anyone enlighten me? Thanks. [...] You can configure sysctl options if you create the file /etc/sysctl.conf and put your changes there. Is this what you're looking for? That happens in the late stages if the game. init is running the main rc script, disks have been fscked and mounted, swap is up. It does happen before network setup, IIRC. It depends on what variables the OP is talking about I guess. DAvid ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: no root login after changing shell
thomas may wrote: Hi, i wanted to change the shell for user root to bash. This is not a good idea. You don't want to use a shell on a partition other than / (e.g. /usr, /usr/local) for root. If those partitions, or shared libraries like libc, get trashed, you are in a world of pain. Much better is to continue to use /bin/sh (or whatever other statically-linked shell takes your fancy) by default. Then, when you want the comfort of bash, all you have to do is run exec bash from the command line and voila! you're using bash. This is how all my servers are set up. 10 keystrokes is a small price to pay for peace of mind. David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: SSH Daemon not running
[EMAIL PROTECTED] wrote: After a standard install of 4.7 the ssh daemon isn't running even though it shows enabled in rc.conf, could it be something else I'm missing. It was my understading that ssh got installed by default. I did go into /etc/ssh/sshd_config and made sure I uncommented the port 22, etc. but still no help. I tried reinstalling just to make sure I didn't accidentally mess something up but the same thing keeps happening. Can someone please point me in the right direction? Does it start if you run /usr/sbin/sshd manually? Are any error messages recorded in /var/log/security or /var/log/messages? David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Large memory issues
Irvine Short wrote: Hey All [snip] and it was slightly better - managed to boot halfway with everything running out of memory. I then found that this: options MAXDSIZ=(2048*1024*1024) options MAXSSIZ=(128*1024*1024) (and also 64MB) options DFLDSIZ=(512*1024*1024) worked fine but not as expected - limit reports datasize unlimited. I've managed to crank it up as far as options MAXDSIZ=(3568*1024*1024) This is for a squid proxy. Right now I have % ps uUnobody USER PID %CPU %MEM VSZ RSS TT STAT STARTED TIME COMMAND nobody 132 0.0 55.6 2185016 2184992 ?? S18Aug03 43:08.02 squid -NsY nobody 133 0.0 0.0 864 412 ?? Ss 18Aug03 0:28.88 (unlinkd) (unlinkd) which is pretty sweet. Which reminds me, I have to tweak my squid conf a bit to let it blow another gig or so of RAM :) I'd kind of like a default of 512MB with the option to loosen it. Any offers as to what's the proper way to do this? All's well at the mo but it doesn't seem right. man limits Configging MAXDSIZ to 3500*1024*1024 did the same as 4095 Also, when the machine boots it says: CPU: Intel(R) Xeon(TM) CPU 2.40GHz (2392.25-MHz 686-class CPU) Origin = GenuineIntel Id = 0xf27 Stepping = 7 Features=0xbfebfbffFPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,CLFLUSH,DTS ,ACPI,MMX,FXSR,SSE,SSE2,SS,HTT,TM,PBE Hyperthreading: 2 logical CPUs real memory = 4025942016 (3931584K bytes) and then later it says on the console something like: 256MB of RAM over 4GB ignored. Seems silly to waste 256MB RAM so any hints would be appreciated here too. How can you address more than 2^32 bytes of RAM with a 32 bit processor? :) Intel have invented a kluge called PAE (Page Address Extensions which lets you eke a bit more out of the current Pentium architecture -- but single processes are still limited to a 4Gb address space. And the talk a while back on freebsd-current was that it has a slight performance impact as well. David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Large memory issues
Irvine Short wrote: [...] and then later it says on the console something like: 256MB of RAM over 4GB ignored. Seems silly to waste 256MB RAM so any hints would be appreciated here too. How can you address more than 2^32 bytes of RAM with a 32 bit processor? :) Yeah, I know about PAE, but it's a 4GB machine so why is there 256MB over the 4GB limit? I dunno, if you count the chips what do you come up with? I recently bought a whole pile of servers (HP DL380s if you've been following my trials and tribulations) that are designed to accept up to 6Gb on the motherboard. I think it's a bit of a gimmick really. One of the servers came bundled with 4.5Gb RAM. We wanted 4Gb, but the machine comes with 512 by default (2x256 chips [1]). Our supplier just stuck in 4x1Gb chips without bothering to remove the existing chips. As the servers all had different chip sizes, 256Mb, 512 and 1Gb, by mixing and matching I was able to remove the wasted 512Mb and install it in another server. Maybe you can do something like that. Or put it on a shelf for spare parts. David [1] DIMMs, SIMMs, whatever they call them these days. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
A DLT unit on /dev/sa0 is really slow
List, I have an external DLT tape (a COMPAQ DLT VS 40/80 5133 according to camcontrol). It seems glacially slow to me. Last night I backed up many gigabytes of files from a remote machine over NFS. It took over five hours to do it. My overworked Sun E-450 will do this in a shade over three hours. What is more, the remote machine and my FreeBSD box have gigabit Ethernet adapters, and both machines, as well as the switch they are connected to all report 1000baseT full duplex. The sun, on the other hand, is running 100baseT. When I run a tar t to list the contents of the archive, the output is really slow. It's taken about half an hour so far to list the first 50 or so files (and none are exceptionally big, 10Mb on average). When I listen to the tape drive, I hear it start and stop frequently, with lots of pauses. I would have expected it to spin continuously. It is, after all, a *streaming* device. All of which makes me think that something is going unspecified somewhere. Something like a preferred block size? 'mt stat' produces the following: Mode Density Blocksize bpi Compression Current: 0x40 variable 0IDRC -available modes- 0:0x40 variable 0IDRC 1:0x40 variable 0IDRC 2:0x40 variable 0IDRC 3:0x40 variable 0IDRC - Current Driver State: at rest. - File Number: 0 Record Number: 0Residual Count 0 Can I manually configure a mode that would have better performance characteristics? Thanks for any tips you might have, David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: National Security Backdoor in telnetd - all versions.
Joshua Oreman wrote: On Mon, Aug 25, 2003 at 11:22:12PM -0400 or thereabouts, Kevin shampoo Nadeau wrote: Hello [ ... ] telnetd is infected with a national security backdoor in all non-source compiled versions of FreeBSD, NetBSD and OpenBSD. If you download the source code for telnetd and compile it to compare the file size of the stock or out of the box version of telnetd versus the source-compiled version - you will clearly see a difference - which is the backdoor. This information should be forwarded to CERN and the BSD and Linux development teams. [ ... ] Please don't feed the troll. The size difference is in stripped vs non-stripped, btw. telnetd eh? how... quaint :) David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Need help with a strange mail/domain problem!
Thomas Beutler wrote: Hi everyone! I'm a total stranger to the Unix world... too long have I dwelled in the Windows area, and now I'm searching for new horizons and worlds to explore... the Unix world. I looked around and found FreeBSD interesting enough to install... and here I am with a strange problem: I installed FreeBSD 5.0-Release and got the system running. So far - so good. Second task: Install bind and get the DNS running. So far - so good - at least I think so. Third task was to get my email work... and now the problems begins. I own the domain beutler.se, and my computers name is visthusboden. The server accepts incoming mail all right - no problems there... BUT... ...whatever I do, all outgoing mail get the address [EMAIL PROTECTED] instead of [EMAIL PROTECTED] What am I doing wrong? You didn't visit http://www.postfix.org/ :) Without wishing to start a flame war, if you are just starting out with smtp on Unix, I would recommend skipping Sendmail using Postfix instead. Having used both (sendmail for years, postfix for months), I wouldn't want to go back to Sendmail even if you paid me. Postfix is a joy to use. I don't think anyone will say that about sendmail. And the configuration files are human-readable. David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Using a RAID-card with FreeBSD
Johan Paul wrote: Hi all, This might be kind of a newbie question - I apologize for that. This is my first time I use a hardware RAID card on a server. I was wondering if there are any issues with FreeBSD (4.8R) regarding these cards? Will FreeBSD see the RAID1 as a single hard drive that I just need to partition and label? And if I need to replace a failed hard Yes, it just sees a single ordinary drive. drive in the RAID array I assume the card will do the rebuilding of the array while FreeBSD will function as normal even when I need to reboot the machine? Yes, the RAIDing is completely transparent to the OS. You don't even have to reboot the machine if the disks are hot-swappable. Just pull the dead one out, install the new one, and the controller will bring it up to speed. RAID-1 is much faster to rebuild than RAID-5 (just a casual observation; I'm not sure if that's a hard-and-fast rule). I checked that the card I intended to use (Promise FastTrak TX2) is supported by FreeBSD so I don't assume any problems installing FreeBSD either. Thanks for any answers :-) you're welcome. David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: max group name length
David Bear wrote:
was just trying to determine the maximum string length of a group
name.
found
struct group {
char*gr_name; /* group name */
but no size.
any pointers (with limits)?
Hmm, I had a browse through the kernel source for a while but didn't
find anything definite. What I do know is that you should endeavour to
keep the length no greater than 8. Up to 15 is probably ok as well on
modern kernels, and group names longer than 15 is getting a bit silly.
Sorry no to have anything more precise.
David
___
[EMAIL PROTECTED] mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-questions
To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: softupdate and squid-cache ?
User Frankb wrote: Hi I'm setting up a Squid cache running on latest 4.8 release and I wondering about disk write performances as it is a crucial point on that kind on machine. My question : does softupdate slow down disk writes ? What you *really* want to do is put the disk cache on a second disk. It's the only game in town for Squid. I mounted a disk as /cache and told squid to use that for its cache, keeping its logs under /var on the first disk. I ran tunefs to give the /cache disk the following characteristics (based on analysis of my first generation squid server): tunefs: average file size: (-f)8192 tunefs: average number of files in a directory: (-s) 256 Be aware though that by default squid will write so many small files that you'll run out of i-nodes before you run out of disk-space. I have to reformat the disk this summer when things are quieter, and I can get away with a smaller disk cache on the system disk. At the moment, with the above tunefs settings, I'm using around 3-4% more available i-nodes than file blocks used as per df -i output. I'm not sure how that translates to specifying the number of i-nodes when I reformat the disk. Hope this helps, David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Eliminating noise from secondary MX
Karl Pielorz wrote: [...] Or, secondly - as was cleverly suggested to me a while ago - setup a 3rd MX that has a IN A PTR to your primary MX, and make it the highest priority... e.g. mx0.mydomain.com PRI 20 mx1.mydomain.com PRI 30 mx2.mydomain.com PRI 40 (Which is really just a different name for mx0) That way, you'll probably find most the spam hits the highest priority MX (which is, in reality your primary MX). I think you mean highest-numbered MX, which is the lowest priority MX, but yeah, I get the picture. Thanks for the tip, I might give it a try. David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Need a little help with a proof of concept cgi.
Dragoncrest wrote: hi all. I've got a little proof of concept idea I want to do and I was looking to get some advice from some of the CGI experts on this list to help me with it. What I want to do is simple. I have like 3 users I want to track and retrieve the following stats on. 1. Total number of total messages received since XX date @ xx time. (since last clearing of counter) 2. Total number of messages filtered as spam. 3. Total number of messages filtered as viruses. I'm sure that there has to be a way to do a very simple cgi script that will be able to be called each time an email is received and after it is sorted so as to increment the individual counters related to each of the sorting criteria. Then when someone loads the cgi script via a Unclear on the concept. I think you want to separate the tallying of stats from the display. The stats can be kept up to date via procmail recipes, content filtering or logfile tailing. That happens all the time. When someone hits your CGI script, you only have to retrieve the statistics, whether they be from a flat file, a DBM file or a database table, and format the results as you see fit. This way you can also produce a test browser that works from the command line. web browser off your web server they can see the stats for each user and clear the counters/mailboxes if necessary. The stats would be my biggest concern. I'm no huge cgi expert and I also figured that this might be useful for a lot of other people in some form so I'm sharing the idea in hopes that those of you with more experience with CGI might be able to come up with a working cgi script. I don't think it will be anything huge, but I'm unsure where to start. Thanks for the input everyone! :) First off, start gathering your stats. That has nothing to do with CGI, and will probably present enough challenges as it is to keep you busy for a while. You might also want to look at packages that do this sort of thing already, like mailgraph. (But otherwise your idea sounds like an excellent learning opportunity). Good luck, DAvid ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: BTX halted when installing on a HP Proliant DL360 G3
David Landgren wrote: Replying to myself, for the sake of the archives. List, I just received a new batch of servers to deploy. Among the lot I have 3 HP Proliant DL360 G3 servers that refuse to run the installation CD. I think this is because HP have switched to Ultra 320 drives and I don't think FreeBSD has caught up with that. I have two 18Gb 15000rpm drives mounted in RAID-1. The BIOS says Proliant System BIOS - P31 (03/01/2003). I've tried various BIOS options, but the boot always goes like this: Miniboot 4.1 Attempting Boot From CD-ROM CD Loader 1.01 Building the boot loader arguments Lookup up /BOOT/LOADER... Found Relocating the loader and the BTX Starting the BTX loader BTX Loader 1.00 BTX version is 1.01 Console: internal video/keyboard BIOS CD is cd0 BIOS drive A: is disk0 inf=000d err= efl=00030006 ... BTX halted I.e.: it looks like it's when it hunts for hard disks. I trawled the archives and learnt that people had similar problems on Compaq hardware, although with Adaptec controllers. The solution there was to create a DOS partition on the disk beforehand. I tried this, with no success. I've tried booting from 4.8-RELEASE, 5.1-RELEASE and the 5.1 miniloader, all with similar results. I also tried booting of an OpenBSD 3.3 CD. It gets a bit further, but when it comes to choosing I for Install, it reports that there are no hard disks available. It seems like FreeBSD simply hasn't caught up yet. I had a look at the Testdrive machines offered by HP, but they're still running 4.8 on a DL360 G2. If anyone has some idea as to how to proceed, I'd be really grateful. The only alternative would be to try and install Linux. But I've been unhappy with Linux performance under load in the past (and it's these servers I'm replacing) so it would be sad to continue in that vein. Thanks for the clues, Daisuke Koike got me onto the right track by pointing me to http://www.freebsd.org/cgi/getmsg.cgi?fetch=61247+63991+/usr/local/www/db/text/2003/freebsd-stable/20030525.freebsd-stable The thread there talks about disabling a BIOS setting named Virtual boot device. Unfortunately my BIOS doesn't have that option. I had another look, and found a different option named Virtual install disk and this one was enabled. I set it to disabled, rebooted, and FreeBSD booted successfully from the CD-ROM. Hope this helps someone in the future. Thanks again Daisuke-san. David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: List of popular DNS blacklists for spam?
Philip Hallstrom wrote: Hi all - I'm testing the use of some DNS blacklists to cut down on my spam and there are a lot of servers to choose from. I was wondering if there is a list of popular servers. I've seen comparison lists, but I'm hoping for more of a survey of sysadmins and what they use (if anything). relays.osirusoft.com -- has a feed from SPEWS, not everyone likes what that represents dnsbl.njabl.org -- very good, although slow to remove sites that fix things up relays.ordb.org -- pretty conservative, no false positives list.dsbl.org -- extremely good, but I get a few false positives from time to time multihop.dsbl.org -- a bit too severe for my tastes blackholes.wirehub.net -- seems ok, but as it's at the end of my list it doesn't have the chance to catch much, I suppose. I've heard favourable reviews of opm.blitzed.org, blackholes.five-ten-sg.com and proxies.relays.monkeys.com Before you start using remote blacklists, you should set up your own restrictions, blocking on garbage HELO, envelope forgeries and the like. David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: How to close port 22 ?
Olivier Nicole wrote: But I strongly suggest that you use SSH instead of telnet to allow remote connection to your machine. uh, maybe he doesn't want any remote connections to his machine? ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Software Watchdog?
Lasse Laursen wrote: Hi, Are there any apps. like 'watchdog' - http://packages.debian.org/stable/admin/watchdog.html - for FreeBSD? I have been searching the ports tree and freshmeat but was unable to locate anything usefull. What are you trying to do? Dan Bernstein's daemontools will look after any number of daemons and make sure that they keep running correctly. You have to buy into djb's mindset though. Some people hate the way it creates directories in / by default. More info can be found at http://cr.yp.to/daemontools.html David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Which Processor and motherboard is better (new to FreeBSD)
Paredes Sánchez Martín A. wrote: Hi: I want to build a new PC with FreeBSD, which processor is best for FreeBSD. Intel builds processors and motherboards, which gave me a feel of good performance between this two elements. Intel has something called Hyper-Threading Technology, Which turbo charges your PC to respond to today's multitasking lifestyle. 4.7 knows about hyper-threading, insofar as the kernel config file admits an 'HTT' option. As to whether it's better or not to have it, I can't really say. 2.4GHz is so ridiculously fast (kernel compiles in less than two minutes, postfix in less than a minute)... I suspect it will be difficult to detect the difference between with/without in everyday use. I'm building a new mail relay, and it turns out the machine was specified as RAID-5. I suspect that that is going to have a much more adverse impact on performance than HTT or not. As to responding to today's multitasking lifestyle, I sooner see PAE implemented in the kernel, to unlock the memory I have above 4Gb. David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Which Processor and motherboard is better (new to FreeBSD)
Paredes Sánchez Martín A. wrote: I found in the intel web site that the Hyper-Threading has this requirements: Hyper-Threading Technology requires a computer system with an Intel Pentium 4 processor at 3.06 GHz or higher, a chipset and BIOS that utilize this technology, and an operating system that includes optimizations for this technology. By the way, what is PAE? PAE eq Page Address Extensions. It allows for 36-bit addressing, which thus lets a 32-bit computer get past the 4Gb addressing limit. They have been around on Intel archictecture for ages in some form or other. I first read about them in DDJ. A search on their website turns up http://x86.ddj.com/articles/2mpages/2mpages.htm It's also the thing that Linus Torvalds blasted a while back. http://www.theinquirer.net/?article=7966 Microsoft seem to be fairly gung-ho about it, searching on Goggle for PAE Intel Page Address Extensions brings up lots of links into their site. David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: perl help
Kenzo wrote:
[resend. I sent this yesterday but it doesn't appear to have turned up
on the list.]
I don't know how to write anything in perl and will eventually learn it.
but I was wondering if anyone would help write a quick perl script for
me.
You really ought to try a different forum. Perl Monks
(http://www.perlmonks.org/) springs to mind.
Basically I want the script to look thru a file for certain words and
cound
how many times it finds the word that comes after.
I have a log file that keeps track of E-mail attachments being send and
received, and I want to be able to do a count of certain attachments.
for example. say I see alot of big this, big that and big nothing
I want to be able to see how many times the word that comes after big
appears in the log file.
so the output would be like this.
this5
that 10
nothing 20
Well that's pretty easy...
#! /usr/bin/perl -w
use strict;
my $target = shift or die No target word given on command line;
my $prev;
my %follow;
while( ) {
chomp;
# break the line into elements
my @word = split;
# deal with the leftover of the previous line
++$follow{$word[0]} if $prev and $prev eq $target;
# walk down the line
$word[$_] eq $target and ++$follow{$word[$_+1]} for 0 .. @word - 2;
# carry over the last word to the next line
$prev = $word[-1];
}
# print out the results
print $_\t$follow{$_}\n for sort keys %follow;
I hope this is not too confusing.
I might say the same of my script :)
thanks.
You're welcome.
David
___
[EMAIL PROTECTED] mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-questions
To unsubscribe, send any mail to [EMAIL PROTECTED]
bge network driver not loaded on a HP (Compaq) Proliant DL380 G3
Hello, I am in the process of installing FreeBSD 4.7 on an HP Proliant DL380 G3. At the moment I am blocked by the fact that the kernel does not recognise the network card (or rather, does not load the bge driver). It is a Broadcom BCM5703X Gigabit Ethernet adapter. (In fact there are two). When I run 'pciconf -lv' it displays the following output (not cutpaste) [EMAIL PROTECTED]:1:0:class=[omitted] vendor = 'Broadcom Corporation' device = 'BCM5703X Gigabit Ethernet' class= network subclass = ethernet and same again more or less for the second adapter at none5. And when I run ifconfig -a, they are not present. The GENERIC kernel doesn't load the driver, and the bge driver is included. I rebuilt a kernel stripping out all the other network adapters and still it does not correctly sense the card. How can I force FreeBSD to recognize the adapters? A second problem, less urgent, is that the machine has 5Gb of RAM, and at boot time the kernel says memory above 4Gb ignored While configuring the new kernel file I reviewed everything in LINT but I didn't see anything obvious. Is it possible to get 4.7 to recognize 4Gb or is this something that can only be done on a more recent version, e.g., 4.8 or 5.0? Thanks for any clues I can use, David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: bge network driver not loaded on a HP (Compaq) Proliant DL380G3
Nicolas Kowalski wrote: David Landgren [EMAIL PROTECTED] writes: I am in the process of installing FreeBSD 4.7 on an HP Proliant DL380 G3. At the moment I am blocked by the fact that the kernel does not recognise the network card (or rather, does not load the bge driver). I faced the same problem with a ML370 G3. My ugly workaround was to plug in this machine another - well recognized - network card. I used an Intel EtherExpressPro10/100 (fxp) for this. Just for closure's sake on the list, here's what I did. As I was loading the 4.8-RC2 iso, playing around with this a bit more I seemed really close to success. In /boot/loader.conf I added miibus_load=YES if_bge_load=YES The drivers load correctly; I can see them when I run 'kldstat'. What I seem to be missing, though, is an explicit way of creating bge0 and bge1. I tried adding ifconfig_bge0=inet a.b.c.d netmask 255.255.255.0 in rc.conf, but it still failed to create the interface. After futzing around reading rc and rc.network, the download finished. I burnt the CD, booted it, and opening the emergency holographic shell on vty4, I saw that bge0 and bge1 existed. I'd rather not have to put an RC2 of anything into production, but I'll keep my fingers crossed. The other alternative was to install redhat 7.3 :o( David ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
