--- Day changed --- Log opened Thu Apr 20 23:59:03 2006 02:00 -!- Marcel [~marcel@jharkema.xs4all.nl] has joined #linode-xenbeta 04:10 < TheFirst> looks like the thrashers died 04:10 < TheFirst> yay 04:40 -!- cherring [~dce932aa@webuser.linode.com] has joined #linode-xenbeta 04:41 < cherring> how are your nodes running boys? 04:48 < TheFirst> Timing buffered disk reads: 2 MB in 117.03 seconds = 17.50 kB/sec 04:48 < TheFirst> not a record low yet 04:50 < cherring> Its really crawling right now, been booting for about 10 mins 04:51 < TheFirst> well when you're io is worse than a floppy... 04:51 < cherring> Im very confused by the hosts behaviour 04:52 < TheFirst> nothing new.... 04:52 < TheFirst> problems are known 04:52 < cherring> what are they? 04:52 < TheFirst> and since the last reboot the problem nodes were thrown in to the priority they were prior so we've had troubles since 04:52 < TheFirst> well to start with the scheduler bites 04:53 < TheFirst> then there's the bugs causing crashes and all that... 04:53 < cherring> If I was a problem node would I be told? 04:53 < TheFirst> you can check yourself if you're thrashing 04:53 < cherring> ow? 04:53 < cherring> *how? 04:53 < TheFirst> check your mem usage 04:54 < TheFirst> and load levels 04:54 < cherring> I have always had swap usage, but when I was a normal node I never lost performance tokens 04:54 < TheFirst> sure...swap isn't bad necc. it's the people actively using the swap that cause problems 04:56 < cherring> well when the load on the host is going well then my load average is always low 04:56 < TheFirst> sure 04:57 < TheFirst> but if your node is idle and you're pulling huge load numbers somethings wrong 04:57 < TheFirst> from what caker was saying earlier thrashers never realize what they're doing 04:57 < cherring> whats huge? 04:57 < TheFirst> sounded like pre io tokens he'd put thrashers in a lower queue that they were still in and had never heard a complaint 04:57 < TheFirst> way i understand it 1 is the equiv of 100% 04:58 < TheFirst> but i may be mistaken on that 04:58 < cherring> hmmmm 04:58 < cherring> I dunno 04:58 < cherring> maybe I will cry 04:59 < TheFirst> well i think caker said next time we have problems he's gonna throw us back on uml until xen gets fixed up 04:59 < cherring> hmmmm 05:00 < cherring> Im sure it'll get worked out 05:00 < TheFirst> http://en.wikipedia.org/wiki/Load_average .. decent explanation of load avgs 05:00 < TheFirst> yah .. just not ready for node use 05:00 < TheFirst> the io scheduler problem and host panics really need to be fixed 05:02 < cherring> yes it does 05:03 < TheFirst> at least caker is proactive about things...if this is a sign of how a linode crowd handles things i'd hate to know how it is at other places 05:03 < TheFirst> linode crowd seems to generally be more aware of the setup and consequences and all that 05:04 < cherring> true, Im quite happy with all that, I hope I get a heads up though if my node is a problem 05:04 < TheFirst> i guess that's one of the most requested features...i brought that up to caker yesterday...email notice or something on problem nodes ... he said it was a long requested feature 05:05 < TheFirst> but then again...on the uml with tokens there wasn't really an issue 05:05 < cherring> yeah 05:06 < TheFirst> course now that it is if i were to be pointed in the direction of thrashers they'd better be track stars cuz i got a bat with their name on it ;) 05:07 < cherring> but they are busy already 05:07 < cherring> lol, well I really hope Im not one of them, maybe should open a ticket 05:07 < mikegrb> lolz 05:07 < cherring> save us mike 05:07 < TheFirst> that's just his stupid script 05:08 < cherring> hmmmm weird 05:08 < TheFirst> another thing that deserves a bat whipping 05:08 < cherring> lol 05:08 < mikegrb> lolz 05:09 < cherring> well when he moved people before, my node was flying 05:13 < cherring> oh well Im gonna jet, hope it settles down 05:13 -!- cherring [~dce932aa@webuser.linode.com] has quit [Quit: CGI:IRC] 05:27 -!- MarcelH [~marcel@jharkema.xs4all.nl] has joined #linode-xenbeta 05:33 -!- Marcel [~marcel@jharkema.xs4all.nl] has quit [Ping timeout: 480 seconds] 05:33 -!- Marcel__ [~marcel@jharkema.xs4all.nl] has joined #linode-xenbeta 05:40 -!- MarcelH [~marcel@jharkema.xs4all.nl] has quit [Ping timeout: 480 seconds] 07:09 -!- sednet [~86920016@webuser.linode.com] has joined #linode-xenbeta 07:10 < sednet> Disk performance seems to be back to near zero on host56. Anyone know whats going on on this host? 07:56 -!- sednet [~86920016@webuser.linode.com] has quit [Quit: CGI:IRC (Ping timeout)] 10:00 -!- brianrumburg [~bri@COX-66-210-194-30.coxinet.net] has joined #linode-xenbeta 10:01 < linbot> New news from forums: Kernel: 2.6.16.1-linode18 with NPTL/TLS support in Linode.com Announcements 10:05 -!- brianrumburg [~bri@COX-66-210-194-30.coxinet.net] has quit [Quit: ] 10:37 -!- sednet [~86920016@webuser.linode.com] has joined #linode-xenbeta 10:55 -!- brianrumburg [~bri@COX-66-210-194-30.coxinet.net] has joined #linode-xenbeta 12:35 -!- Marcel__ [~marcel@jharkema.xs4all.nl] has quit [Quit: Leaving] 12:35 -!- Marcel [~marcel@jharkema.xs4all.nl] has joined #linode-xenbeta 12:57 < taupehat> hrm 12:58 * taupehat suspects the trasher is back 12:58 < taupehat> err, thrash^Wyeah, trasher... whoever it is is trashing the host! 13:00 < TheFirst> you're right 13:01 < TheFirst> but it's not THAT bad yet 13:01 < TheFirst> Timing buffered disk reads: 54 MB in 3.01 seconds = 17.97 MB/sec 13:04 < taupehat> it's just getting choppy 13:04 * taupehat kind of misses io_tokens 13:05 < TheFirst> hah i was thinking the same thing 13:05 < TheFirst> i bet caker does too :P 14:57 < valen2> Sounds like a good movie tagline... "Thrash Master 5000 ... The thrash is back!" 15:04 < brianrumburg> TheFirst: Where are you getting your MB/sec measurement? 15:07 < valen2> I think it's... 15:07 < valen2> hdparm -t /dev/hda1 15:08 < valen2> Replace hda1 with your main virtual image. 15:08 < brianrumburg> thanks valen 15:10 -!- Breaker_1 [~Breaker_1@165.138.28.199] has joined #linode-xenbeta 15:11 < Breaker_1> anyone else on host56 having problems at the moment? 15:12 < valen2> hdparm timing locked up on me. LOL 15:12 < mikegrb> lolz 15:13 < valen2> And apparenly my Celery (Celeron 400A) is faster than the server now... Timing buffered disk reads: 66 MB in 3.06 seconds = 21.58 MB/sec 15:14 -!- TheFirst [gaveup@your.friendly.neighborhood.hellmouth.info] has quit [Ping timeout: 480 seconds] 15:14 -!- taupehat [me@taupehat.com] has quit [Ping timeout: 480 seconds] 15:15 < Breaker_1> where are we at right now, far as the beta is concerned? how long before we've got more stability? 15:15 < Breaker_1> lol 15:15 < mikegrb> lolz 15:19 < brianrumburg> Wow. Could it be that I've got that running in a loop right now? (c; 15:21 -!- taupehat_ [~c6ed8603@webuser.linode.com] has joined #linode-xenbeta 15:21 < taupehat_> host56 is bewm 15:21 < taupehat_> ping caker 15:21 < brianrumburg> I think I'm gonna start trying to thrash the hell out of the disk just for sport. ... or maybe a fork bomb in a boot script. (http://en.wikipedia.org/wiki/Fork_bomb). That would be fun. 15:22 < Breaker_1> I'd be happy with finally finishing setting up my svn server and getting some of my work done 15:27 < brianrumburg> cool, I can't ping my node, now. 15:28 < brianrumburg> JobID: 246163 - Host initiated restart 15:28 < Breaker_1> I still can't 15:28 < brianrumburg> check your Host's Job Queue 15:29 < Breaker_1> I got host initiated restart 15:29 < Breaker_1> status failed 15:29 < brianrumburg> yea 15:29 < brianrumburg> same 15:29 < Breaker_1> I always get that though when the host goes down like this 15:29 < brianrumburg> I'm gonna be pissed if someone actually DID fork bomb the server... 15:29 < Breaker_1> I don't think you can really be pissed 15:29 < Breaker_1> this is a beta server 15:29 < Breaker_1> I'm jsut frustrated 15:30 < brianrumburg> true true. 15:30 < brianrumburg> Not really pissed just seems abusive if that were the case. 15:30 < Breaker_1> yeah 15:31 < Breaker_1> I think the nodes are isolated well enough that a forkbomb inside of a node wouldn't hurt the others though 15:31 < Breaker_1> it'd have to be done on the server itself 15:31 < Breaker_1> I'm just used to the vmware esx server 15:32 < brianrumburg> Maybe, but I'm not going to pretend to understand how Xen passes process back to the host kernel, either. 15:32 < Breaker_1> I trialed Xen and Esx 15:32 < brianrumburg> yea? 15:32 -!- snorp [~snorp@CPE-65-28-73-215.kc.res.rr.com] has joined #linode-xenbeta 15:32 * snorp chuckles at topic 15:32 < Breaker_1> I do understand why we're so unstable 15:32 < Breaker_1> Xen requires more configuring 15:32 < brianrumburg> Is there a public ESX host? 15:32 < Breaker_1> no 15:32 < Breaker_1> and if there were it'd be expensive 15:33 < Breaker_1> a 4cpu license for esx (academic pricing) is $6000 15:33 < Breaker_1> for just one server 15:33 < brianrumburg> High licensing. 15:33 < brianrumburg> lol 15:33 < mikegrb> lolz 15:33 < Breaker_1> and that's just to use esx on your server 15:33 < Breaker_1> not the smp license 15:33 < Breaker_1> or the license for their managment application 15:33 < brianrumburg> I think this could be considered academic, right? We're all learning stuff here. 15:34 < Breaker_1> conversely... it's vmware 15:34 -!- bendy24 [~scott@bendy.netop.oftc.net] has joined #linode-xenbeta 15:34 < Breaker_1> so you can run anything you want 15:34 < Breaker_1> and not have to worry about file systems or compatilibty 15:34 < Breaker_1> our esx test bed we have right now we've got a small cluster running linux, freebsd, novell, and (hehe) dos 15:34 < Breaker_1> oh and server 2003 15:35 < Breaker_1> trying to talk my boss into using it 15:35 < Breaker_1> over all it'd cut our costs down 15:35 < brianrumburg> We haven't taken the ESX plunge here, yet. 15:35 < Breaker_1> even though it's expensive as hell 15:35 < brianrumburg> still working on Server. 15:35 < brianrumburg> atop 2003 15:35 < brianrumburg> free as hell. 15:36 < Breaker_1> right now, here .. we have no virtualization at all 15:36 < Breaker_1> none 15:36 < Breaker_1> and we have 3 racks full of servers at 1 to 2 % utilization 15:36 < brianrumburg> DOH 15:36 < Breaker_1> yeah 15:36 < Breaker_1> my boss is just... slow 15:36 < Breaker_1> and old 15:36 < Breaker_1> he doesn't want to move away from having 1 app per server 15:36 < brianrumburg> we've recently started building custom XP embeded images for use on our VM's 15:36 < Breaker_1> because he doesn't really understand it 15:37 < Breaker_1> and he's scared it'll crash and burn and everyone will hate him 15:37 < linbot> New news from forums: Is host56 down? in Xen Public Beta 15:37 < Breaker_1> New news from the channel: Yes, host56 is down. 15:37 < Breaker_1> and isn't new news redundant? 15:38 -!- sednet [~86920016@webuser.linode.com] has quit [Quit: CGI:IRC (Session timeout)] 15:38 < brianrumburg> Had similar resistance here, for a bit. Till we were really able to sit down and show it was better and definately no worse as a point of failure. 15:38 < Breaker_1> yeah... I tried 15:38 < brianrumburg> lol 15:38 < mikegrb> lolz 15:38 < Breaker_1> I don't know if I didn't explain it well 15:38 < Breaker_1> or if my boss just really wasn't listening 15:38 < Breaker_1> but he didn't understand it at all it seemed to me 15:38 < brianrumburg> prolly the latter. 15:38 < brianrumburg> (c; 15:39 < Breaker_1> he understood all we need is a san and 4 or 5 nice servers and we're set 15:39 < Breaker_1> that we can ditch all our old servers 15:39 < Breaker_1> and the electricity bills 15:39 < gpd> what about the opposite of virutalization: clustering - is there anything coming along that turns 10 boxes into 1 uber box without the need to rewrite all the apps? 15:39 < Breaker_1> and the noise 15:39 < mikegrb> Linodes are comming back up 15:39 < brianrumburg> thanks mike 15:39 < Breaker_1> I hate working in a server room 15:39 < Breaker_1> it's cold and loud 15:40 < Breaker_1> even with my headphones on I can hear the noise 15:40 < brianrumburg> you should look into noise cancelling headphones... that is if you can't convice the boss about virtualization. 15:40 < gpd> pah - i have 2x -80C freezers, centrifuges, 1x HPLC, 1x Fridge iwth old compressor... 15:40 < gpd> and NO WINDOW! 15:40 < Breaker_1> yeah I have the no window 15:40 < brianrumburg> yeek. 15:41 < Breaker_1> everything else I lack 15:41 < Breaker_1> especially because I don't know what a HPLC is 15:41 < brianrumburg> lol 15:41 < mikegrb> lolz 15:41 < gpd> it's a machine that goes 'bing' 15:41 < Breaker_1> YOu need that 15:41 < Breaker_1> what about the machine we lease from the company we sold it to 15:41 < Breaker_1> do we have that? 15:41 < brianrumburg> so, like a phone, only different noise? 15:41 < Breaker_1> or is that the machine that goes ping? 15:41 < snorp> lol 15:41 < mikegrb> lolz 15:41 < snorp> lol 15:41 < mikegrb> lolz 15:41 < Breaker_1> (that was a monty python joke) 15:42 < snorp> hmm. 15:42 < Breaker_1> ahhh.. the meaning of life 15:42 < Breaker_1> that's one of my favorite sections 15:42 < Breaker_1> because that really is how hospitals are 15:42 < Breaker_1> we're missing something... 15:42 < Breaker_1> missing something... ... 15:42 < Breaker_1> the patient! 15:42 < Breaker_1> where is she? 15:43 < gpd> openMOSIX is the closest to my ^^ afaik 15:47 -!- iggy [~iggy@12.45.184.235] has joined #linode-xenbeta 15:50 < iggy> when did host56 reboot last? 15:51 < gpd> 12:39 mikegrb> Linodes are comming back up 15:51 < iggy> mine didn't automatically boot back up 15:52 < iggy> what tz you in? 15:52 < brianrumburg> yea, I'm still down as well... 15:52 < taupehat_> mine neither, had to issue a boot 15:52 < taupehat_> it's up now 15:52 < brianrumburg> mkay 15:52 < gpd> iggy: Pacific 15:52 < gpd> ie. about 10 minutes ago 15:53 -!- taupehat [me@taupehat.com] has joined #linode-xenbeta 15:53 < taupehat> whoo 15:53 -!- taupehat_ [~c6ed8603@webuser.linode.com] has quit [Quit: CGI:IRC 0.5.7 (2005/06/19)] 15:54 < iggy> can one unenroll from the xenbeta? 15:55 < taupehat> iggy: put in a support ticket 15:55 < iggy> k, danke 15:55 < taupehat> bitte 15:56 < taupehat> heh 15:57 < gpd> Das Computermachine ist nicht fuer gefingerpoken und mittengrabben. Ist easy schappen der shpringenverk, blowenfusen unt poppencorken mit spitzensparken. Ist night fuer gewerken bei die Dummkopfen. Das rubbernecken Sichtseeren keepen das cotton-pickenen Hands in die Pockets muss; relaxen und watschen die Blinkenlichten. 15:57 < taupehat> hosts booting are thrashing it again (at least, I hope that's what is causing the slagging) 15:57 < taupehat> gpd: hah, and I'm listening to kraftwerk right now. And your mock-german is sucky =P 15:57 < gpd> sorry - any german wrt computers reminds me of that :) 15:57 < gpd> not mine - that was posted in a room i worked in in 1995 15:57 < gpd> it is the original blinkenlights 16:00 < taupehat> ahh 16:05 -!- TheFirst_ [~GaveUp@CPE-70-92-72-102.new.res.rr.com] has joined #linode-xenbeta 16:05 < TheFirst_> what happened 16:06 < Breaker_1> You guys planning on running xen on freebsd? and making a freebsd host? 16:06 < Breaker_1> lol 16:06 < mikegrb> lolz 16:24 -!- marc_in_lux [~gergesm@cable-83.217.130.243.coditel.net] has joined #linode-xenbeta 16:26 < marc_in_lux> hello 16:26 < marc_in_lux> my linode's powered off. about 4 hours ago 16:26 < snorp> host56 went down 16:26 < snorp> it's back now 16:26 < marc_in_lux> is there anything wrong or can i just go and start it. 16:27 < marc_in_lux> no automatic starting for the linodes? 16:27 < snorp> mine started automatically, but maybe yours had a problem? 16:27 < Spads> http://www.helenakeeffe.com/archives/ZoeBaker.mp3 <-- Sir Butts-a-Lot 16:27 < marc_in_lux> I found nothing in the queue... last entry was from April 19. 16:30 < iggy> marc_in_lux: yeah, apparently, a few of us didn't get automatically restarted 16:30 -!- TheFirst [gaveup@your.friendly.neighborhood.hellmouth.info] has joined #linode-xenbeta 16:39 -!- bendy24 [~scott@bendy.netop.oftc.net] has quit [Quit: leaving] 16:40 -!- Breaker_1 [~Breaker_1@165.138.28.199] has quit [Quit: Leaving] 16:52 -!- lucca [~lucca@export.accela.net] has joined #linode-xenbeta 16:52 < lucca> aha 17:32 -!- marc_in_lux [~gergesm@cable-83.217.130.243.coditel.net] has quit [Quit: Leaving] 17:35 < linbot> New news from forums: IPSet in Feature Request/Bug Report 17:59 -!- brianrumburg [~bri@COX-66-210-194-30.coxinet.net] has left #linode-xenbeta [] 19:43 -!- snorp is now known as snorp|out 21:14 -!- lucca [~lucca@export.accela.net] has quit [Quit: restarting emacs] 21:15 -!- lucca [~lucca@export.accela.net] has joined #linode-xenbeta --- Log closed Fri Apr 21 23:59:01 2006