Re: System resource monitoring and reporting utility ?
I've worked with tons of programs that do this - none of them have filled all my requirements. One thing, though - I didin't want to rely on SNMP - My ultimate solution would be a system that will ssh to a machine and run commands, process the output. In no particular order: 1. Monit is good for status - not history. If you want to know what's running NOW, or get en email when disk usage goes above 80% - things that can be checked remotely (ping, HTTP/S, Mysql) then monit is great. No pretty graphs, though. 2. Nagios is the best for all-around monitoring - but configuration is a pain. Installing the Groundwork framework makes this a cinch, though installing groundwork itself is slightly painful and pretty invasive on the host machine. If you have a machine to dedicate to this, this is probably the best solution for most cases. 3. mrtg is good, especially if all your devices speak SNMP. It's cute an simple - you cron a perl script to run every, say, 5 minutes. It outputs a bunch of graphs. I don't think it does warnings/emails, etc. 4. Orcallator. Some large companies I know use this. I tried it once - I seem to remeber that I liked it, though I couldn't online any of the features that I thought I liked :-) 5. Zabbix - Very robust, but requires their agents to be installed on all non SMTP machines. If you are willing to do this, then this is a Great option. Really - Zabbix is a mature system that works great. 6. Zenoss - compatible with nagios plugins. I would say this solution reminds me of a mach-up of zabbix and nagios. 7. Cacti - Non recommended. I don't remember why. On 8/9/07, Shachar Shemesh [EMAIL PROTECTED] wrote: Oren Held wrote: A friend of mine (Amnon) found Munin (http://munin.projects.linpro.no/), which is a great system resource grapher tool which has plugins for almost everything from swap, ntp time drifts, disk temperature - to mysql queries per second. However, it draws graphs, I'm not sure it can send alerts. You can set limits (like highest cpu temperature or free disk space) which tag the whole node as red, maybe it's even capable of notifying.. worth a check I guess. I don't know about alerts, but I find that it is the best tool I know for getting the general health of a system. That is something no graph specific test can tell you, because it often involves measurements you did not think of before they happened. For anyone who is interested in seeing it in action, Hamakor's new server has it running, open for all to see: http://hamakor.org.il/munin/ Shachar = To unsubscribe, send mail to [EMAIL PROTECTED] with the word unsubscribe in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED] = To unsubscribe, send mail to [EMAIL PROTECTED] with the word unsubscribe in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: System resource monitoring and reporting utility ?
Oh - I forgot. Nagios let me do my whole check-through-ssh thing. meaning: I put the nagios-plugins somewhere on the remote machines and create a nagios user on each. Configure key authentication for this user on all the machines. Then I configured nagios to check_by_ssh all the things I wanted checked. Configuration would have killed me had I not used the Groundwork Framework. On 8/9/07, Mike Tewner [EMAIL PROTECTED] wrote: I've worked with tons of programs that do this - none of them have filled all my requirements. One thing, though - I didin't want to rely on SNMP - My ultimate solution would be a system that will ssh to a machine and run commands, process the output. In no particular order: 1. Monit is good for status - not history. If you want to know what's running NOW, or get en email when disk usage goes above 80% - things that can be checked remotely (ping, HTTP/S, Mysql) then monit is great. No pretty graphs, though. 2. Nagios is the best for all-around monitoring - but configuration is a pain. Installing the Groundwork framework makes this a cinch, though installing groundwork itself is slightly painful and pretty invasive on the host machine. If you have a machine to dedicate to this, this is probably the best solution for most cases. 3. mrtg is good, especially if all your devices speak SNMP. It's cute an simple - you cron a perl script to run every, say, 5 minutes. It outputs a bunch of graphs. I don't think it does warnings/emails, etc. 4. Orcallator. Some large companies I know use this. I tried it once - I seem to remeber that I liked it, though I couldn't online any of the features that I thought I liked :-) 5. Zabbix - Very robust, but requires their agents to be installed on all non SMTP machines. If you are willing to do this, then this is a Great option. Really - Zabbix is a mature system that works great. 6. Zenoss - compatible with nagios plugins. I would say this solution reminds me of a mach-up of zabbix and nagios. 7. Cacti - Non recommended. I don't remember why. On 8/9/07, Shachar Shemesh [EMAIL PROTECTED] wrote: Oren Held wrote: A friend of mine (Amnon) found Munin (http://munin.projects.linpro.no/), which is a great system resource grapher tool which has plugins for almost everything from swap, ntp time drifts, disk temperature - to mysql queries per second. However, it draws graphs, I'm not sure it can send alerts. You can set limits (like highest cpu temperature or free disk space) which tag the whole node as red, maybe it's even capable of notifying.. worth a check I guess. I don't know about alerts, but I find that it is the best tool I know for getting the general health of a system. That is something no graph specific test can tell you, because it often involves measurements you did not think of before they happened. For anyone who is interested in seeing it in action, Hamakor's new server has it running, open for all to see: http://hamakor.org.il/munin/ Shachar = To unsubscribe, send mail to [EMAIL PROTECTED] with the word unsubscribe in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED] = To unsubscribe, send mail to [EMAIL PROTECTED] with the word unsubscribe in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Help with Linux on sparc64
On 7/31/07, Rami Rosen [EMAIL PROTECTED] wrote: Hello Nadav, What binary distribution has the best support for Sun machines? The first distribution which was equipped with support for Sun SPARC64 is Ubuntu; this happened after some attempts to boot Fedora Core (by Dave Miller and others) failed giving panic. This of course does not imply the Ubuntu is the best one , but it is the most veteran in this field. So I suggest you try it (you can try Ubuntu live-cd first to see what are the results). That's a very dangerous statement - I've installed at least 3 linux distros on my 64-bit ultra5 years before Ubuntu was created. Granted, one of them was debian :-) but after boot the keyboard stops working (It only has USB keyboardmouse). to make sure this keyboard is identified please try : tail -f /var/log/messages, and then unplug the USB keyboard and plug it back in. Do you see any messages ? Regards, Rami Rosen On 7/26/07, Nadav Shemer [EMAIL PROTECTED] wrote: Hi. I have a Sun Ultra 45 and I've tried (unsuccessfully) to install some kind of Linux distribution on it. I've tried Debian netinst (both stable and testing) and Gentoo (stage2) CDs, but after boot the keyboard stops working (It only has USB keyboardmouse). Booting from the serial console, It hangs with the following message: su: Cannot register IRQ 1 Has anyone had any luck with this particular machine? What binary distribution has the best support for Sun machines? Can anyone help me? Nadav Shemer Tehuti Networks [EMAIL PROTECTED] = To unsubscribe, send mail to [EMAIL PROTECTED] with the word unsubscribe in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Help with Linux on sparc64
I installed linux on my ultra5 quite a bit ago. Have you cheked out ultralinux, or better yet, auroralinux.org? -mike On 7/27/07, Nadav Shemer [EMAIL PROTECTED] wrote: Thank you for your comments. I'll definitely try stage 3. On 7/27/07, Geoffrey S. Mendelson [EMAIL PROTECTED] wrote: Before you attempt to get Linux to run on it, it would be a good idea to make sure there is not a hardware problem. I started the process with a machine running Solaris. I cleared out a slice, intending on installing linux on it. Nadav
Re: Hardware Recommendation
I'd also be wary of computers stores in Israel - A customer was screaming at the store people - it seems they switched out his Micron memory for Generic brand. The store just said it's the same thing - you won't notice the difference On 7/17/07, Shahar Dag [EMAIL PROTECTED] wrote: Hi You should also consider the quality of the service you can get in Israel. from rommers HP aren't so good from personal experience IBM tends to be petty (for example if you have a tiny crack in the case, they may refuse to fix a problem in the computer unless you will also replace the case) have fun Shahar - Original Message - From: Oded Arbel [EMAIL PROTECTED] To: linux-il@cs.huji.ac.il Sent: Sunday, July 15, 2007 6:21 PM Subject: Hardware Recommendation Hi All. I'm in the market for a new laptop, and I hope that people on the list can share their experience with current laptops, specifically in regards to Linux compatibility and reliability. I'm interested in spending less money, but still get a brand with good (preferably world-wide) support - otherwise I'd go with a good Thinkpad which I know from experience is a great product, so currently I'm looking at these brands: HP Pavillon Compaq Presario Dell (whatever - can't make heads or tails of their brand names) Also a couple related questions: - Did anyone have any success in getting a brand laptop in Israel without paying the MS-Windows tax ? - Did anyone manage to get a Dell with Ubuntu installed ? From talking to some Israeli sales reps that distribute Dell laptops I got mostly huh?!?s. Thanks in advance -- Oded = To unsubscribe, send mail to [EMAIL PROTECTED] with the word unsubscribe in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED] = To unsubscribe, send mail to [EMAIL PROTECTED] with the word unsubscribe in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Hardware Recommendation
Be sure to check out http://consumerist.com/consumer/insiders/22-confessions-of-a-former-dell-sales-manager-268831.phpif you're in the market for a dell. -mike On 7/15/07, Oded Arbel [EMAIL PROTECTED] wrote: On Sun, 2007-07-15 at 18:11 +0200, ASAF HALILI wrote: about the ibm thinkpad, all of the series known as great laptops, comfortable to use and very massive computers. if you have the budget, go for it, about the linux support i'm not sure. I've used several Thinkpad models and all support Linux very well, including support for the special keys, bluetooth, wireless and possibly even the fingerprint reader (haven't tried it and it isn't bundled yet with major distros). Didn't have any problem at all in installation and day to day usage, though the Intel GMA versions might be a bit problematic if you don't use the most recent X.org version with the new (and experimental) intel mode setting driver. Also I wasn't very successful with using 3D acceleration with the ATI versions - even with ATI's binary driver. = To unsubscribe, send mail to [EMAIL PROTECTED] with the word unsubscribe in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: [SUMMARY] Current status of Israeli ISPs
In the US, I worked for a local computer store - it basically had all the Jewish Community's business - Anyway, one room of the store was an ISP - It was 2 racks - one for servers, the other for network. Server rack had COTS desktops running BSD (I think) - 2 * (mail, DNS, RADIUS,News) servers It was dial-up internet, so he had (I guess) a leased line from a telephone bank off-site. My point is that the whole operation was one room - those 2 racks and 3 desks covering phone support and sys admins. ISP'ing, at least in the US *can* be done on a smallish scale - on the order of a few hundred customers. On 7/9/07, Amos Shapira [EMAIL PROTECTED] wrote: On 09/07/07, Geoffrey S. Mendelson [EMAIL PROTECTED] wrote: On Sun, Jul 08, 2007 at 08:01:50PM +0300, David Smith wrote: If I understand correctly, in order to set up an small ISP, the following infrastructure would be needed: A connection to at least one bigger ISP, and preferably also to IIX. A server to handle DNS/mail/accounting etc and possibly routing Optional: a dedicated hardware router An 'ISP' connection to Bezeq and Hot. You forgot an ISP license from whatever ministry issues them. This requires a lot of money. The cheapest way to do it is to buy an existing licensed ISP. That's how BEZEQ BENLUMI did it, they bought ISDNNET which got their license by buying a previous ISP. I know Orange was refused a license, I don't know how they eventually got one. I didn't quite follow the entire discussion from top to bottom, but is there an option to become a reseller? That way you get the basic infrastructure from an approved ISP but create your own brand and support lines. Not sure it makes sense, especially in the Israeli context, but apparently it's common in other parts of the world. --Amos
Re: 32Gb servers?
Have you looked into using memcached ? http://www.danga.com/memcached/ : memcached is a high-performance, distributed memory object caching system, generic in nature, but intended for use in speeding up dynamic web applications by alleviating database load. Danga Interactive http://www.danga.com/ developed memcached to enhance the speed of LiveJournal.com http://www.livejournal.com/, a site which was already doing 20 million+ dynamic page views per day for 1 million users with a bunch of webservers and a bunch of database servers. memcacheddropped the database load to almost nothing, yielding faster page load times for users, better resource utilization, and faster access to the databases on a memcache miss. On 6/19/07, Ariel Biener [EMAIL PROTECTED] wrote: On Monday, 18 בJune 2007 10:32, Amos Shapira wrote: Amos, what is your budget for this ? There are a few options: SGI 1200 or SGI 2100 HP DL360G5 Dell PowerEdge 1950 IBM x3455 IBM x3550 No name Intel based boards stuff (if you're talking about 5 exits, try Data-Store - they are one and the same AFAIK - www.datastore.co.il) However, for a server with 32GB memory, I would go for one of the brand names. --Ariel Hello, Where would you go if you had to get a 32Gb RAM server, much preferably rack-mounted. Don't care so much about CPU or very fast disks, just needs lots of RAM. Can run either Windows (possibly developer's preference) or Linux (my preference). Dell's smallest server which supports 32Gb jumps the price to over 45k when it comes with 32Gb, 44k of this is just for the ram. I'm trying to dig the other big name brands (IBM, Sun, SGI, HP) but so far their web sites weren't very helpful to understand what can they offer that answers these simple requirements. Thanks, --Amos = To unsubscribe, send mail to [EMAIL PROTECTED] with the word unsubscribe in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED] -- -- Ariel Biener *.il EFnet Admin PGP: http://www.tau.ac.il/~ariel/pgp.html -- -- Ariel Biener e-mail: [EMAIL PROTECTED] PGP: http://www.tau.ac.il/~ariel/pgp.html To unsubscribe, send mail to [EMAIL PROTECTED] with the word unsubscribe in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: 32Gb servers?
You can probably buy a small cluster (4 or 8 machines?) for 45k. Then, use a shared memory system (ccNUMA maybe?) On 6/18/07, Amos Shapira [EMAIL PROTECTED] wrote: Hello, Where would you go if you had to get a 32Gb RAM server, much preferably rack-mounted. Don't care so much about CPU or very fast disks, just needs lots of RAM. Can run either Windows (possibly developer's preference) or Linux (my preference). Dell's smallest server which supports 32Gb jumps the price to over 45k when it comes with 32Gb, 44k of this is just for the ram. I'm trying to dig the other big name brands (IBM, Sun, SGI, HP) but so far their web sites weren't very helpful to understand what can they offer that answers these simple requirements. Thanks, --Amos = To unsubscribe, send mail to [EMAIL PROTECTED] with the word unsubscribe in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Solid state disks (was 32gb computers)
How do solid-state disks compare in speed to DDR-RAM? On 6/18/07, Geoffrey S. Mendelson [EMAIL PROTECTED] wrote: Wouldn't a memory based solid state disk do better for a lot less money? You can even get them with multiple interfaces so they can be shared between servers. Geoff, -- Geoffrey S. Mendelson, Jerusalem, Israel [EMAIL PROTECTED] N3OWJ/4X1GM IL Voice: (07)-7424-1667 U.S. Voice: 1-215-821-1838 Visit my 'blog at http://geoffstechno.livejournal.com/ = To unsubscribe, send mail to [EMAIL PROTECTED] with the word unsubscribe in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: 32Gb servers?
Good one. On 6/18/07, Gilad Ben-Yossef [EMAIL PROTECTED] wrote: Amos Shapira wrote: The local VAR is 5 Exists, alhough being Oz side right now I reckon this won't help you much Uhhh, 5 Exists, in Ramat Hachayal? I remember finding something with them that I couldn't find anywhere else (probably some specific hardware recommended by MAV). Yeap, those ones. May I ask what sort of need you have it for? In our case our developers want a very highly accessible very large database to just be stored in memory (e.g. for a lookup while a customer of our clients clicks on an HTML page link). We want to host the world championship of Core War. :-) Gilad -- Gilad Ben-Yossef [EMAIL PROTECTED] Codefidence. A name you can trust(tm) http://www.codefidence.com Phone: +972.3.7515563 ext. 201 | Cellular: +972.52.8260388 SIP: [EMAIL PROTECTED] | Fax: +972.3.7515503 Lacking fins or tail the gefilte fish swims with great difficulty. -- A Jewish Haiku = To unsubscribe, send mail to [EMAIL PROTECTED] with the word unsubscribe in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Is pthread_mutex_lock fair?
I'm a complete novice when it comes to these things - but is it possible that somewhere, something is doing an optimization? Compiler? CPU? Perhaps some subsystem along the way sees that fair scheduling would take much longer. -mike On 6/10/07, Muli Ben-Yehuda [EMAIL PROTECTED] wrote: On Sun, Jun 10, 2007 at 10:06:40AM +0300, Gilad Ben-Yossef wrote: AFAIK sched_yeild() precious meaning is and has always been: put me at the end of the run queue for my priority. If there is no one else with my priority, I'll run again. Quote akpm: Changed sched_yield() semantics. sched_yield() has changed dramatically, and it can seriously impact existing applications. A testcase (this is on 2.5.46, UP, no preempt). http://lkml.org/lkml/2002/12/2/155. By calling sched_yielf(), the scheduler we re-arrange tun run queue so that (assuming all worker threads are of the same priority), the thread releasing the lock will be put at the back of the run queue and some other threads, which ahs been marked ready to run after the lock has been released will get the chance to take the lock, that's why I expect sched_yield() to do the job. Note that for an SMP machine things are rather more complicated and also if there are numeroud tasks in different priroities... :-( Yes, you are correct. sched_yield() will give an appearance of fairness. But consider the cost (see below) - IMO it changes the semantics of the test program in such a way as to make it meaningless. Shahar, let us know what happens :-) sched_yield() makes it run 20 times longer for TEST_SIZE=1000, and the difference grows with the test size. moren:/home/muli # unset DO_SCHED_YIELD; time taskset 01 ./mf 5 Thread 40800940 started as thread #0 Thread 41001940 started as thread #1 Thread 41802940 started as thread #2 Thread 42003940 started as thread #3 Thread 42804940 started as thread #4 Results: 234626 1177504 1155320 2187538 5245012 Joined thread #0 Joined thread #1 Joined thread #2 Joined thread #3 Joined thread #4 real0m1.187s user0m1.000s sys 0m0.116s moren:/home/muli # export DO_SCHED_YIELD=1; time taskset 01 ./mf 5 using sched_yield() Thread 40800940 started as thread #0 Thread 41001940 started as thread #1 Thread 41802940 started as thread #2 Thread 42003940 started as thread #3 Thread 42804940 started as thread #4 Results: 106710 106711 106712 106711 106709 Results: 213147 213147 213148 213147 213147 Results: 319311 319312 319313 319312 319311 Results: 425650 425652 425653 425652 425651 Results: 531269 531270 531269 531269 531269 Results: 637252 637254 637253 637252 637253 Results: 743678 743681 743680 743679 743679 Results: 849960 849963 849962 849961 849961 Results: 957552 957554 957554 957552 957553 Results: 1065060 1065063 1065064 1065061 1065061 Results: 1172691 1172694 1172695 1172692 1172691 Results: 1280342 1280345 1280345 1280342 1280342 Results: 1388211 1388214 1388214 1388211 1388211 Results: 1495994 1495998 1495999 1495995 1495994 Results: 1603876 1603880 1603881 1603878 1603877 Results: 1711619 1711623 1711624 1711622 1711619 Results: 1819511 1819513 1819515 1819513 1819510 Results: 1927503 1927506 1927507 1927505 1927503 Results: 199 201 203 200 197 Joined thread #0 Joined thread #1 Joined thread #2 Joined thread #3 Joined thread #4 real0m19.249s user0m3.008s sys 0m15.909s #include stdio.h #include stdlib.h #include pthread.h #include unistd.h #define TEST_SIZE 1000 volatile int results[TEST_SIZE], position=0; static int do_sched_yield = 0; int worker( int threadnum ) { static pthread_mutex_t mutex=PTHREAD_MUTEX_INITIALIZER; printf(Thread %lx started as thread #%d\n, pthread_self(), threadnum); pthread_mutex_lock(mutex); while( positionTEST_SIZE ) { // printf(results[%d]=%d\n, position, threadnum); results[position++]=threadnum; pthread_mutex_unlock(mutex); if (do_sched_yield) sched_yield(); pthread_mutex_lock(mutex); } pthread_mutex_unlock(mutex); return 0; } int *histo; void printres(int numthreads) { int i; printf(Results: ); for( i=0; inumthreads; ++i ) printf(%d , histo[i]); printf(\n); } int main(int argc, char * argv[] ) { int numthreads=atoi(argv[1]); if( argc2 || numthreads==0 ) exit(1); pthread_t *threads=calloc(sizeof(pthread_t), numthreads); histo=calloc(sizeof(int), numthreads); int i; int oldpos=position; if (getenv(DO_SCHED_YIELD)) { do_sched_yield = 1; printf(using sched_yield()\n); } for( i=0; inumthreads; ++i ) { pthread_create(threads+i, NULL, (void *(*)(void *))worker, (void *)(unsigned long)i); } /* Reap the results every second */ while( oldposTEST_SIZE ) { sleep(1); for( ; oldposposition; ++oldpos ) { histo[results[oldpos]]++; } printres(numthreads); } for( i=0; inumthreads; ++i ) { pthread_join( threads[i], NULL ); printf(Joined thread #%d\n,