Back to Home / #linode / 2006 / 08 / Prev Day | Next Day
#linode IRC Logs for 2006-08-22

---Logopened Tue Aug 22 00:00:05 2006
00:00|-|kokoko [] has joined #linode
00:08|-|konoko [] has quit [Ping timeout: 480 seconds]
00:09|-|kokoko [] has quit [Ping timeout: 480 seconds]
00:13<erikh>taupehat: that tattoo you pasted in that other channel is positively evil
00:20<taupehat>ain't it? But I'm not pasting it here!
00:21<TheFirst>wtf you talkin' about
00:34|-|Olrick [] has quit [Ping timeout: 480 seconds]
00:37|-|Olrick [] has joined #linode
00:55|-|internat [] has joined #linode
01:02|-|internat [] has quit [Quit: This computer has gone to sleep]
01:02|-|kokoko [] has joined #linode
01:04|-|tekno [~rac@] has joined #linode
01:04<tekno>i'd like to load the vncservice at boot, before users log on w/ debian, how can i do this? there doesn't seem to be a /etc/rc.d/rc.local file...
01:04<tekno>any hints
01:05|-|tekno [~rac@] has left #linode []
01:30|-|Spads [] has joined #linode
01:57|-|internat [] has joined #linode
01:59|-|spr [] has quit [Quit: Spoon!]
02:04|-|womble [] has quit [Quit: What's behind the round window...]
02:20|-|Spads [] has quit [Ping timeout: 480 seconds]
03:31|-|Spads [~crack@] has joined #linode
04:12|-|okokok [] has joined #linode
04:18|-|konoko [~Vampire@] has joined #linode
04:19|-|kokoko [] has quit [Ping timeout: 480 seconds]
04:24|-|okokok [] has quit [Ping timeout: 480 seconds]
04:31|-|afv-13 [~afv-13@] has joined #linode
04:31|-|afv-13 [~afv-13@] has quit []
05:21|-|chrisw [~chrisw@] has joined #linode
05:22|-|chrisw [~chrisw@] has quit []
06:09|-|sec39 [] has quit [Ping timeout: 480 seconds]
06:29|-|sec39 [] has joined #linode
06:45|-|linville [] has joined #linode
07:56|-|konoko [~Vampire@] has quit [Ping timeout: 480 seconds]
08:03|-|ximbiot [] has joined #linode
08:19|-|javafueled [] has joined #linode
08:19|-|javafueled [] has quit []
08:19|-|javafueled [] has joined #linode
08:29|-|NeonNero [~nn@] has joined #linode
08:34|-|jungleg [] has joined #linode
08:38|-|Dreamr3 [] has joined #linode
08:44|-|Dreameer3 [] has quit [Ping timeout: 480 seconds]
09:05|-|okokok [] has joined #linode
09:18|-|dejavu [] has joined #linode
09:20|-|dejavu1 [] has joined #linode
09:21<dejavu1>Can someone give me some hints on this problem "VFS: Cannot open root device" in UML???
09:22|-|dejavu1 [] has quit []
09:24|-|dejavu [] has quit [Quit: CGI:IRC (Ping timeout)]
09:30|-|NeonNero [~nn@] has quit [Quit: Everybody who believes in telekinesis, raise my hand!]
09:45<rickl>in the lish console, I'm getting the error message:
09:45<rickl>__alloc_pages: 0-order allocation failed (gfp=0x1d2/0)
09:48<npmr>means you're out of memory
09:49<rickl>hrmm strange
09:49<rickl>i had 50megs free like 5 minutes ago
09:50<rickl>maybe... it's all the visitors to a gallery2 site
09:50<rickl>or webmin
09:50<kvandivo>maybe you could run 'top' and find out
09:50<rickl>cant run anything
09:51<rickl>server is at a standstill
09:51<rickl>grep -E -q (^|:)/sbin($|:)
09:51<rickl>is running
09:51<rickl>... hack attempt?
09:51<kvandivo>ssh io_status
09:51<rickl>yeah i'm out of io tokens
09:52<rickl>this literally happened in a matter of minutes
09:52<rickl>i have 20 apache processes running
09:52<kvandivo>as soon as you hit swap, the io tokens disappear into the ether
09:52<rickl>what about that grep command?
09:52<rickl>that doesn't look good...
09:52<npmr>it's grep
09:53<npmr>you know what grep is, right?
09:53<rickl>but... i didnt run it
09:53<bendy24>kill it
09:53<rickl>i wish i could
09:53<rickl>*laughs out loud*
09:53<rickl>where's caker or mikegrb
09:53<npmr>what's its parent process?
09:53<rickl>can't tell... i just have a few-minute-old ps listing
09:53<npmr>try 'ps fwaux'
09:54<rickl>nothing is going through...
09:54<rickl>not even from lish console
09:54<npmr>is this a linode 100?
09:54<npmr>and you've got 20 apache processes going?
09:54<rickl>just recently
09:55<npmr>lots of visitors?
09:55<kvandivo>tune? why tune? just throw it out there and see what happens
09:55<rickl>i guess so... maybe i set my apache thread stuff too low
09:55<rickl>and its starting too many processes
09:55<npmr>it surely sounds like it
09:56<rickl>gallery2 probably takes up a lot of ram...
09:56<bendy24>it is a hog
09:56<npmr>gallery2 is a big ol' php lump
09:56<kvandivo>ands bendy would know
09:56<rickl>last time i do my boss a favor
09:56<bendy24>rickl: i've limited apache to 4 threads
09:56<bendy24>on a 100
09:56<rickl>and put a gallery website up on my server for 60 slacking employees
09:57<rickl>can you send me your config line?
09:57<rickl>i'm interested to see what you have
09:57<bendy24>one sec
09:57<npmr>StartServers 2
09:57<npmr>MinSpareServers 1
09:57<npmr>MaxSpareServers 3
09:57<npmr>MaxClients 5
09:57<npmr>MaxRequestsPerChild 15
09:57<rickl>and how come after logging into lish i wont get a bash prompt
09:57<npmr>^-- on my 200
09:57<npmr>rickl, because lish is your shell on the host system, not bash
09:58<rickl>yeah i know... i guess its just too thrashy to spawn a bash process
09:58<rickl>ok i give up... rebooting server
09:59|-|javafueled [] has quit [Remote host closed the connection]
09:59<kvandivo>i use 1, 1, 3, 9, 2000 for those same npmr settings
10:00<internat>ooc is there a good reason to actually have a small max requests per child?
10:00<internat>assuming that apache isnt leaking memory?
10:00<rickl>im beginning to think its better to have a higher MRPC
10:01<rickl>my apologies to anyone on host41
10:02<npmr>internat, it causes the apache master process to kill off a few extra idle children
10:02<kvandivo>internat: ya.. that's why i have it at 2k.. i figure that's enough that i won't notice the startup costs
10:02<npmr>keeps it closer to MinSpareServers than MaxSpareServers
10:03<internat>why? to me that doesnt make sense, when the servers are idle it should be killing them off anyway, maxrequestsperchild shouldnt matter
10:03<rickl>i have IfModule prefork.c and worker.c
10:03<npmr>internat, i have a year's worth of munin data that says otherwise, so bugger off
10:04<rickl>oh whoops
10:04<internat>im not having a go at you, im trying to understand how that works
10:04<rickl>i had MaxClients at 100
10:04<kvandivo>internat: and totally ignore the comments in the httpd.conf file that the authors of apache wrote to help you. npmr clearly knows better
10:05<Spads>npmr: are you on prefork?
10:05<npmr>i am
10:05[~]Spads is trying worker out
10:05<npmr>last i knew php still requires it
10:05<Spads>oh yeah php
10:05[~]Spads uninstalled php
10:05<Spads>good riddance
10:05<internat>doesnt php5 support the new one now?
10:05<npmr>yeah, i'd like to
10:05<npmr>i've got users stuck on squirrelmail, though
10:06<internat>i use horde myself, which i imagine will break when i change to php5 so i guess ill be waiting a while anyway
10:06<npmr>horde is also written in php?
10:06<rickl>is worker enabled by default?
10:07<rickl>horde is php
10:07<internat>sure is
10:07|-|javafueled [] has joined #linode
10:07<npmr>StartServers 2
10:07<npmr>MinSpareServers 1
10:07<npmr>MaxSpareServers 3
10:07<npmr>MaxClients 5
10:07<npmr>MaxRequestsPerChild 15
10:07<npmr>damn mousey
10:07<rickl>wrong one? hehe
10:07|-|javafueled [] has quit [Remote host closed the connection]
10:07<rickl>what about for worker?
10:08<rickl>or is that not enabled by default...
10:08|-|javafueled [] has joined #linode
10:08<npmr>i think worker was the default when i installed apache2
10:08<npmr>but i have it set to prefork
10:08<rickl>what config settings do you have for worker?
10:09<npmr>i'm using prefork
10:09<npmr>i have prefork tuned, not worker
10:09<npmr>my worker settings are moot
10:09<rickl>and its strange that MaxClients in prefork is max number of processes
10:09<npmr>they are probably the debian defaults
10:09<rickl>when in worker, it's "maximum number of simultaneous client connections"
10:09<rickl>does it still mean max number of processes?
10:10<npmr>MaxClients means the same thing in both prefork and worker
10:10<rickl>lousy config file
10:10<npmr>it's the maximum number of simultaneous client connections
10:10<rickl>it says otherwise
10:10<npmr>do you know what prefork and worker are?
10:11<rickl>i have some idea, but no idea which one I'm using
10:11<npmr>prefork is old school apache
10:11<npmr>it's how apache 1.3 worked *all the time*
10:11<npmr>and it's one process per client connection
10:12<npmr>therefore MaxClients = maximum simultaneous connections = maximum number of processes
10:12<npmr>worker is newfangled multithreaded apache, only available in 2.0 and up
10:13<npmr>worker can do multiple client connections per process
10:13<rickl>ah... it means the same thing in the clients aspect, but not the processes aspect
10:13<rickl>and on apache2 i guess i'm running worker by default
10:13<npmr>so: MaxClients = maximum simultaneous connections > maximum number of processes
10:14<rickl>hrmm seems i'm using prefork
10:16<rickl>so with MaxClients = 5, does that mean you can only have 5 "users" connecting to the server?
10:16<npmr>it means that there can be up to five simultaneous connections
10:17<rickl>so... no more than 5 people viewing a site on the server at once.
10:17<bendy24>rickl: you may also want to revert to gallery1 which seems to run alot faster
10:17<npmr>generally, when a web browser is done downloading an object, it closes the connection
10:17<rickl>or, instead of using gallery... maybe i'll just make a static page
10:20<bendy24>2.2 was supposed to be on par with 1.x, but it hasnt appeared yet
10:20|-|internat [] has quit [Quit: This computer has gone to sleep]
10:23<TheFirst>if you dont want comments try minimal gallery...
10:26<rickl>ok i think my problem was a misconfiguration of apache
10:26<rickl>cause people are looking at the gallery now and things seem fine (for now)
10:26<TheFirst>if you dont need apache specific things consider'll save a lot of resources
10:32<rickl>yeah i'm pretty stuck on apache
10:32<rickl>thanks guys
11:09<@caker> <-- har
11:11<taupehat>that's funny though
11:28|-|Spads [~crack@] has quit [Ping timeout: 480 seconds]
11:28<@linbot>New news from forums: Immediate timeout, even though message says timed out after in General Discussion <>
11:54|-|Dreamr3 [] has quit [Ping timeout: 480 seconds]
11:57|-|spr [] has joined #linode
11:58|-|jekil [] has joined #linode
12:13|-|afv-13 [] has joined #linode
12:16|-|Dreamer3 [] has joined #linode
12:46|-|javafueled [] has left #linode []
13:13|-|Battousai [] has joined #linode
13:29|-|afv-13 [] has quit [Ping timeout: 480 seconds]
13:34|-|jungleg [] has quit [Quit: CGI:IRC (Session timeout)]
14:06|-|Spads [] has joined #linode
14:11|-|spr_ [~spr@] has joined #linode
14:12|-|spr [] has quit [Ping timeout: 480 seconds]
14:15|-|M-guy [] has quit [Ping timeout: 480 seconds]
14:15|-|linville [] has quit [Quit: Leaving]
14:35|-|claviola [] has quit [Quit: LOLZ @ NOT DEFAULT IRSSI QUIT MASSAGE]
14:46|-|M-guy [] has joined #linode
15:16|-|spr__ [] has joined #linode
15:17|-|spr__ changed nick to spr
15:24|-|spr_ [~spr@] has quit [Ping timeout: 480 seconds]
15:32|-|hellon [~rvn@] has joined #linode
15:44[~]warewolf grub ... savedefault --default=1 --once to reboot his box into windows at h ome
15:45|-|hellon [~rvn@] has left #linode []
17:25|-|rko [] has joined #linode
17:51|-|spr [] has quit [Quit: "Ah, the bounce has gone from his bungie"]
17:56<taupehat>holy wow
18:21|-|internat [] has joined #linode
18:38|-|sonorous [] has quit [Quit: reconnecting]
18:39|-|sonorous [] has joined #linode
18:43|-|internat [] has quit [Quit: This computer has gone to sleep]
18:46<@caker>taupehat: mommie!
18:55[~]TheFirst calls the whitecoats on caker
18:59|-|internat [] has joined #linode
18:59|-|internat [] has quit [Remote host closed the connection]
19:01|-|internat [] has joined #linode
19:11|-|internat [] has quit [Quit: This computer has gone to sleep]
19:23|-|rickl [] has quit [Quit: meow]
19:37|-|internat [] has joined #linode
19:38|-|funkycow [] has joined #linode
19:40|-|funkycow changed nick to rick
19:45|-|rko [] has quit [Ping timeout: 480 seconds]
19:47|-|rick [] has quit [Quit: meow]
19:47|-|funkycow [] has joined #linode
19:47|-|funkycow changed nick to rick
19:48|-|rick [] has quit []
19:54|-|funkycow [] has joined #linode
19:54|-|funkycow changed nick to rick
20:01|-|internat [] has quit [Quit: This computer has gone to sleep]
20:12|-|jekil [] has quit [Ping timeout: 480 seconds]
20:12|-|internat [] has joined #linode
20:17|-|GetzU [~Tango@] has joined #linode
20:17<GetzU>hello caker
20:18<GetzU>hello mikegrb
20:23<GetzU>any guys from linode team out here
21:06|-|GetzU [~Tango@] has quit [Ping timeout: 480 seconds]
21:45|-|GetzU [~Tango@] has joined #linode
21:46|-|Dreamer3 [] has quit [Quit: Leaving]
21:58|-|TheFirst [] has quit [Ping timeout: 480 seconds]
21:59|-|taupehat [] has quit [Ping timeout: 480 seconds]
22:00[~]caker calls HE
22:03|-|Tangoa [~Tango@] has joined #linode
22:04<@caker>Tangoa: what's up?
22:04<Tangoa>i need xtra ip's
22:04<Tangoa>for my linode200
22:05<Tangoa>but no response from the linodeteam to my ticket
22:05<@caker>Tangoa: ticket number?
22:05<Tangoa>let me check
22:05<Tangoa>one minute\
22:06<Tangoa>and 21519
22:06<Tangoa>please look into it
22:07<Tangoa>coz i have many irc servers running
22:07<@caker>Tangoa: I'm taking care of another issue at the moment, but I'll look into it once I'm done with this
22:07<Tangoa>so tough to get u online
22:08<Tangoa>different time zone
22:08|-|GetzU [~Tango@] has quit [Ping timeout: 480 seconds]
22:08<Tangoa>and why cant i log into this room with my nick Tango
22:09<Tangoa>check this
22:09<Tangoa>Tango Nickname is already in use.
22:09<@caker>Tangoa: /msg nickserv help
22:09|-|internat [] has quit [Quit: This computer has gone to sleep]
22:10<Tangoa>u guys are not planning in selling real dedicated servers
22:11<@caker>Tangoa: nope, just virtual
22:12|-|Dreamer3 [] has joined #linode
22:15|-|taupehat [] has joined #linode
22:19|-|GaveUp [] has joined #linode
22:22<@mikegrb>Tangoa: I've responded to your tickets
22:23<GaveUp>mikegrb: what's up with 56?
22:24<@mikegrb>there were problems with a switch but those are resolved
22:25<GaveUp>hmmm can't connect to my node but can connect to lish...od
22:26<encode>i have a curly question - can you think of any way i could sync time in a network which is locked down really tight, with udp 123 closed
22:26<iggy>there's some http time sync program I saw once
22:27<encode>i wonder how reliable that would be
22:27<encode>considering http would be proxied
22:27<iggy>probably not much
22:27<encode>i'll do some searching
22:27<encode>its got to be better than nothing
22:27<JasonF>encode: ntpdate -u
22:28<JasonF>ntpdate -u
22:28<JasonF>(-u = use unprivledged port)
22:29<encode>JasonF: if only it were that easy...i have a choice of windows 2003 or as/400 systems to use as the internal authoritative server
22:29<@mikegrb>that's just for the local port
22:29<@mikegrb>still uses 123 for the remote port
22:29<JasonF>ah, i misunderstood his ? then
22:31<JasonF>caker: did you guys finally fix the mad reboots of my host?
22:32<JasonF>I've had 43 days of uptime.
22:32|-|Tangoa [~Tango@] has quit []
22:33|-|GaveUp [] has quit [Quit: You're a bloody puppet!]
22:37|-|TheFirst [] has joined #linode
22:59|-|VS_ChanLog [] has left #linode [Rotating Logs]
22:59|-|VS_ChanLog [] has joined #linode
23:06<encode>well, i found a pogram that syncs over http, but it doesnt have any settings for proxy authentication
23:11<iggy>they make usb radio time receivers
23:18<encode>they cost money ;)
23:49|-|Netsplit <-> quits: schultmc, @linbot, insanum, dc0e, WonderClown, rick, jcm, iggy, Olrick, sec39, (+5 more, use /NETSPLIT to show all of them)
23:49|-|Netsplit over, joins: taupehat
23:49|-|Netsplit over, joins: rick
23:49|-|Netsplit over, joins: Spads
23:49|-|Netsplit over, joins: sec39
23:49|-|Netsplit over, joins: Olrick
23:49|-|Netsplit over, joins: WonderClown
23:49|-|Netsplit over, joins: insanum
23:49|-|Netsplit over, joins: jcm
23:49|-|Netsplit over, joins: dc0e
23:49|-|Netsplit over, joins: xptek
23:49|-|Netsplit over, joins: iggy
23:49|-|Netsplit over, joins: linbot
23:49|-|Netsplit over, joins: cout
23:49|-|Netsplit over, joins: Newsome
23:49|-|Netsplit over, joins: schultmc
23:49|-|NetHack #linode [+o linbot] by []
---Logclosed Wed Aug 23 00:00:06 2006