From rnovak at indyramp.com Mon Aug 13 09:08:54 2012 From: rnovak at indyramp.com (Robert Novak) Date: Mon, 13 Aug 2012 09:08:54 -0700 Subject: [Baylisa] BayLISA in San Francisco this week - important notes in case you're going Message-ID: Hi all, Just a reminder of our first meeting outside Silicon Valley in quite a while. BayLISA is hosted by Joyent this month, at their offices in One Embarcadero Center in San Francisco. Details are available at http://www.meetup.com/BayLISA/events/42780152/. There are some things this month that are different from our usual meeting, so I wanted to give you some quick notes on those differences. 1. The meeting is in San Francisco. Might seem obvious, but... yeah. You can take public transportation (see http://tripplanner.transit.511.org/) from the South Bay or East Bay fairly easily, although it will cost you a few bucks if you don't already have a transit pass. I'll probably be taking Caltrain from Palo Alto to Millbrae, and then BART to the Embarcadero station in SF, but you can also take Caltrain to 4th and King and then take MUNI in to downtown/Market Street. Parking at One Embarcadero Center is $3/hr. You can drive up, and you may wish to carpool. 2. You'll need a real name to get in Our hosts are in a shared commercial office facility, and building security at Embarcadero Center requires first and last names for all attendees to ensure admission to the facility. Anyone not on the list may not be allowed in. If you have a fake or partial name on Meetup, or if you have +1s RSVPed, please go back to the site and put your first and last name (and those of your guests, if applicable) into your RSVP comments or your actual name on Meetup. (Note that Bill vs William is close enough, but Unixgod Thundertrousers vs Brian Wilson is not. ) The RSVP list will close at 4pm this Wednesday in order for us to get the list to security in time. 3. There's limited space. We have the list set at 75 people, which is about the capacity of our meeting room. Note that there is a waiting list on Meetup and as people change their plans there may be more seats opened up. People may cancel at any time so keep your eyes open for spaces to open up. over the next couple of days. Don't worry, we'll be back on Silicon Valley turf next month, and most likely for the rest of the year. We're just trying this out as a way to expand our reach and our speaker pool. Depending on the reception we might aim for 2-3 San Francisco or Peninsula meetups a year to accommodate the growing north-of-Stanford sysadmin community. And of course we welcome your suggestions for locations, speakers, topics, sponsors, and so forth as we go forward and continue to grow BayLISA. Thanks, Robert Novak BayLISA President From rayw at rayw.net Tue Aug 14 13:13:37 2012 From: rayw at rayw.net (Ray Wong) Date: Tue, 14 Aug 2012 13:13:37 -0700 Subject: [Baylisa] tomcat and other monitoring? Message-ID: Hey all, not like there's a lot of list traffic/discussions these days, but iirc that's actually one of the accepted uses for the list, so I thought I'd try it out again... Just wondering if other folks are still having to brew up monitoring/graphing solutions. detecting outages and failures seems a given, but there doesn't seem to be much going on WRT graphing trends and such... In the case of monitoring tomcat, I found a lone thread on the cacti forums, which generate some vaguely useful graphs, but the solution is all wrapped up in a rather obfuscating XML-template system that doesn't seem to lend itself well to certain customizations such as running multiple instances, etc. Anyone found a better solution, or figure out enough jmxproxy to come up with more modular, basic scripts to create something a little more flexible and extensible? There seem to be newer graphing options (graphite, carbon, etc), but doesn't seem to be much in the way of actually talking to data sources still. Anyone? Seems like we all used to deal with a lot of this as sysadmins, and maybe my googlefu is not strong, but seems like most of the efforts are all years out of date now. -R> From hso at nosneros.net Tue Aug 14 13:54:57 2012 From: hso at nosneros.net (Holt Sorenson) Date: Tue, 14 Aug 2012 20:54:57 +0000 Subject: [Baylisa] tomcat and other monitoring? In-Reply-To: References: Message-ID: <20120814205457.GB4931@nosneros.net> ohhai, Ray! Fancing meeting you here. :P Graphite/Ganglia are popular these days for monitoring/graphing: http://graphite.wikidot.com/ http://neopatel.blogspot.com/2011/04/logging-to-graphite-monitoring-tool.html http://ganglia.sourceforge.net/ http://blog.sflow.com/2011/12/using-ganglia-to-monitor-java-virtual.html JMX glue for jvm (Tomcat too) -> Ganglia/Graphite: https://github.com/lookfirst/jmxtrans/wiki HTH On Tue, Aug 14, 2012 at 01:13:37PM -0700, Ray Wong wrote: >Hey all, not like there's a lot of list traffic/discussions these >days, but iirc that's actually one of the accepted uses for the list, >so I thought I'd try it out again... > >Just wondering if other folks are still having to brew up >monitoring/graphing solutions. detecting outages and failures seems a >given, but there doesn't seem to be much going on WRT graphing trends >and such... In the case of monitoring tomcat, I found a lone thread on >the cacti forums, which generate some vaguely useful graphs, but the >solution is all wrapped up in a rather obfuscating XML-template system >that doesn't seem to lend itself well to certain customizations such >as running multiple instances, etc. Anyone found a better solution, or >figure out enough jmxproxy to come up with more modular, basic scripts >to create something a little more flexible and extensible? There seem >to be newer graphing options (graphite, carbon, etc), but doesn't seem >to be much in the way of actually talking to data sources still. > >Anyone? Seems like we all used to deal with a lot of this as >sysadmins, and maybe my googlefu is not strong, but seems like most of >the efforts are all years out of date now. -- Holt Sorenson hso at nosneros.net www.nosneros.net/hso From brendon at netcal.com Tue Aug 14 14:04:57 2012 From: brendon at netcal.com (Brendon Baumgartner) Date: Tue, 14 Aug 2012 21:04:57 +0000 Subject: [Baylisa] tomcat and other monitoring? In-Reply-To: <20120814205457.GB4931@nosneros.net> References: <20120814205457.GB4931@nosneros.net> Message-ID: http://www.zabbix.com -----Original Message----- From: Baylisa [mailto:baylisa-bounces at baylisa.org] On Behalf Of Holt Sorenson Sent: Tuesday, August 14, 2012 13:55 To: Ray Wong Cc: baylisa at baylisa.org Subject: Re: [Baylisa] tomcat and other monitoring? ohhai, Ray! Fancing meeting you here. :P Graphite/Ganglia are popular these days for monitoring/graphing: http://graphite.wikidot.com/ http://neopatel.blogspot.com/2011/04/logging-to-graphite-monitoring-tool.html http://ganglia.sourceforge.net/ http://blog.sflow.com/2011/12/using-ganglia-to-monitor-java-virtual.html JMX glue for jvm (Tomcat too) -> Ganglia/Graphite: https://github.com/lookfirst/jmxtrans/wiki HTH On Tue, Aug 14, 2012 at 01:13:37PM -0700, Ray Wong wrote: >Hey all, not like there's a lot of list traffic/discussions these days, >but iirc that's actually one of the accepted uses for the list, so I >thought I'd try it out again... > >Just wondering if other folks are still having to brew up >monitoring/graphing solutions. detecting outages and failures seems a >given, but there doesn't seem to be much going on WRT graphing trends >and such... In the case of monitoring tomcat, I found a lone thread on >the cacti forums, which generate some vaguely useful graphs, but the >solution is all wrapped up in a rather obfuscating XML-template system >that doesn't seem to lend itself well to certain customizations such as >running multiple instances, etc. Anyone found a better solution, or >figure out enough jmxproxy to come up with more modular, basic scripts >to create something a little more flexible and extensible? There seem >to be newer graphing options (graphite, carbon, etc), but doesn't seem >to be much in the way of actually talking to data sources still. > >Anyone? Seems like we all used to deal with a lot of this as sysadmins, >and maybe my googlefu is not strong, but seems like most of the efforts >are all years out of date now. -- Holt Sorenson hso at nosneros.net www.nosneros.net/hso _______________________________________________ Baylisa mailing list Baylisa at baylisa.org http://www.baylisa.org/mailman/listinfo/baylisa From guy at extragalactic.net Tue Aug 14 21:36:09 2012 From: guy at extragalactic.net (Guy B. Purcell) Date: Tue, 14 Aug 2012 21:36:09 -0700 Subject: [Baylisa] tomcat and other monitoring? In-Reply-To: References: Message-ID: <0FC76F80-12DC-410A-8651-A3334826E441@extragalactic.net> Zenoss does both pretty well. Also, there's a Nagios widget that will store the polling results in RRDs, IIRC--may even tie into the UI for effectively built-in graphs. -Guy On Aug 14, 2012, at 1:13 PM, Ray Wong wrote: > Hey all, not like there's a lot of list traffic/discussions these > days, but iirc that's actually one of the accepted uses for the list, > so I thought I'd try it out again... > > Just wondering if other folks are still having to brew up > monitoring/graphing solutions. detecting outages and failures seems a > given, but there doesn't seem to be much going on WRT graphing trends > and such... In the case of monitoring tomcat, I found a lone thread on > the cacti forums, which generate some vaguely useful graphs, but the > solution is all wrapped up in a rather obfuscating XML-template system > that doesn't seem to lend itself well to certain customizations such > as running multiple instances, etc. Anyone found a better solution, or > figure out enough jmxproxy to come up with more modular, basic scripts > to create something a little more flexible and extensible? There seem > to be newer graphing options (graphite, carbon, etc), but doesn't seem > to be much in the way of actually talking to data sources still. > > Anyone? Seems like we all used to deal with a lot of this as > sysadmins, and maybe my googlefu is not strong, but seems like most of > the efforts are all years out of date now. > > -R> > _______________________________________________ > Baylisa mailing list > Baylisa at baylisa.org > http://www.baylisa.org/mailman/listinfo/baylisa From guy at extragalactic.net Tue Aug 28 18:54:27 2012 From: guy at extragalactic.net (Guy B. Purcell) Date: Tue, 28 Aug 2012 18:54:27 -0700 Subject: [Baylisa] Colo space? Message-ID: Hi All, I currently have a 1U server colo'd at HE. I'd like to replace it with a vastly more efficient 2U one I've built, but it requires more juice than I'm contracted for. So, before I go and renegotiate my contract, I thought I'd ask if any one of y'all either has 2U you'd like to rent long-term, or would like to go in with me on a 1/4 cab rental someplace. Anyone? -Guy From pozar at lns.com Tue Aug 28 18:53:57 2012 From: pozar at lns.com (Tim Pozar) Date: Tue, 28 Aug 2012 18:53:57 -0700 Subject: [Baylisa] Colo space? In-Reply-To: References: Message-ID: <7dab08f2-a433-4cf2-981b-befe3c5258bf@email.android.com> Have you talked to the folks at Cernio colo co-op? They rent by the U at Layer42 in Mountain View and United Layer in San Francisco. Tim -- Sent from my phone. Please excuse my brevity. "Guy B. Purcell" wrote: Hi All, I currently have a 1U server colo'd at HE. I'd like to replace it with a vastly more efficient 2U one I've built, but it requires more juice than I'm contracted for. So, before I go and renegotiate my contract, I thought I'd ask if any one of y'all either has 2U you'd like to rent long-term, or would like to go in with me on a 1/4 cab rental someplace. Anyone? -Guy _____________________________________________ Baylisa mailing list Baylisa at baylisa.org http://www.baylisa.org/mailman/listinfo/baylisa From eric at explosive.net Tue Aug 28 19:33:29 2012 From: eric at explosive.net (Eric Sorenson) Date: Tue, 28 Aug 2012 19:33:29 -0700 Subject: [Baylisa] Colo space? In-Reply-To: References: Message-ID: <56E6B376-C6A7-4624-9377-2758BD2FF311@explosive.net> Strongly suggest you talk to Luke Crawford at prgmr.com -- super clueful/responsive and does this kind of thing as bread and butter. -=Eric On Aug 28, 2012, at 6:54 PM, "Guy B. Purcell" wrote: > Hi All, > > I currently have a 1U server colo'd at HE. I'd like to replace it with a vastly more efficient 2U one I've built, but it requires more juice than I'm contracted for. So, before I go and renegotiate my contract, I thought I'd ask if any one of y'all either has 2U you'd like to rent long-term, or would like to go in with me on a 1/4 cab rental someplace. Anyone? > > -Guy > > _______________________________________________ > Baylisa mailing list > Baylisa at baylisa.org > http://www.baylisa.org/mailman/listinfo/baylisa