Back to Home / #linode / 2006 / 12 / Prev Day | Next Day
#linode IRC Logs for 2006-12-11

---Logopened Mon Dec 11 00:00:17 2006
00:24|-|avid [] has joined #linode
00:35|-|avid [] has quit [Quit: avid]
00:40|-|avid [] has joined #linode
00:49|-|avid [] has quit [Quit: avid]
01:08|-|konoko [] has quit [Ping timeout: 480 seconds]
01:11|-|konoko [~Vampire@] has joined #linode
01:54|-|SpaceHobo [~spacehobo@] has joined #linode
02:45|-|andrew_j_w [] has joined #linode
02:50|-|spr [] has quit [Quit: spr]
04:25|-|SpaceHobo [~spacehobo@] has quit [Read error: Operation timed out]
04:26|-|irgeek [] has quit [Quit: irgeek]
04:39|-|tsp [] has quit [Quit: leaving]
04:45|-|konoko [~Vampire@] has quit [Ping timeout: 480 seconds]
05:04|-|konoko [] has joined #linode
05:35|-|SpaceHobo [~spacehobo@] has joined #linode
06:27|-|linville [] has joined #linode
06:53|-|Nigel_ [] has joined #linode
06:57|-|Nigel [] has quit [Ping timeout: 480 seconds]
06:59|-|Angel_Boy [] has joined #linode
07:05|-|Eman [] has quit [Read error: Operation timed out]
07:45|-|Newsome [] has joined #linode
09:17|-|_andrew_j_w [] has joined #linode
09:21|-|andrew_j_w [] has quit [Ping timeout: 480 seconds]
09:31|-|agraman [] has joined #linode
09:36|-|agraman [] has quit [Quit: CGI:IRC (Ping timeout)]
09:37|-|agraman [] has joined #linode
09:43|-|pete [] has joined #linode
09:43|-|agraman [] has quit [Quit: CGI:IRC (Ping timeout)]
09:47|-|agraman [] has joined #linode
09:48|-|pete [] has quit [Quit: CGI:IRC (Ping timeout)]
09:48|-|Eman [] has joined #linode
09:49|-|agraman [] has quit []
09:49|-|agraman [] has joined #linode
09:50<agraman>Is anybody able to tell me what is happening with host44?
09:51<agraman>Perhaps I should have firsyly asked if anyone else is having problems on host44
09:53|-|agraman [] has quit []
09:53|-|agraman [] has joined #linode
09:53|-|agraman [] has quit []
10:23|-|Dreamr_3 [] has joined #linode
10:27|-|spr [] has joined #linode
10:29|-|Dreamr3 [] has quit [Ping timeout: 480 seconds]
11:28|-|afv-13 [] has joined #linode
11:50|-|anderiv [] has joined #linode
11:52|-|heidi [] has quit [Remote host closed the connection]
11:52|-|heidi [] has joined #linode
11:54|-|concha [] has joined #linode
11:56|-|SpaceHobo [~spacehobo@] has quit [Ping timeout: 480 seconds]
11:59|-|Angel_Boy [] has quit [Quit: CGI:IRC (Session timeout)]
12:05|-|smccoy [] has joined #linode
12:05|-|smccoy changed nick to tag
12:05<anderiv>anyone else seeing extreme packet loss to TP currently?
12:06|-|petert [] has joined #linode
12:07|-|FireSlash [] has joined #linode
12:08<concha>yes, very bad
12:08<npmr>not from sbc dsl in chicago, charter cable in michigan, speakeasy in michigan, or speakeasy in los angeles
12:09<npmr>although i am seeing lag between my linode-based irc client and oftc
12:09<anderiv>a traceroute shows that traffice is fine until it hits
12:09<npmr>i'm connected to oxygen
12:11<tag>my packetloss is averaging 35-40% from seattle (through sprintlink)
12:11<anderiv>tag: ditto....from comcast in Mpls.
12:11<petert>i'm interested in reinstalling my linode, and i'd like to migrate the data between images locally if possible. would it be supported for me to pay for extra disk space for one month, then drop it once i'm migrated?
12:12<tag>I'm on a pretty major backbone, I'm extremely confident the problem isn't over here
12:12<tag>or I'd be seeing ops alerts about it like crazy
12:12<npmr>petert, you can do that, but you probably wouldn't have to
12:12<tag>petert: incase you
12:13<npmr>petert, you could just resize your existing images so that they're just barely big enough for your data and deploy a new system in your free space
12:14<petert>npmr: with parted or similar, or is there a linode tool i should use?
12:14<npmr>there's a linode tool
12:14<petert>excellent, thanks. i'll look into that
12:15|-|rainkid [] has joined #linode
12:17|-|jblack [] has joined #linode
12:18<petert>oof, it's going to be tight :)
12:20<rainkid>sooo... whats the issue?
12:22<jblack>I'm having a tough time reaching my node.
12:23<rainkid>bandwidth issue... DoS attack?
12:23<afv-13>we all are
12:23<jblack>every once in awhile I can squeeze in and get a load avearge. Load averages are low, at .30
12:24<rainkid>caker around?
12:24<anderiv>doesn't look like DoS to me...
12:24<rainkid>leeching user?
12:24<jblack>Not specific to my machine, I don't think so.
12:25<anderiv>when pings come back, they're coming back with pretty low latency...typically when I've seen DoS attacks, the routers' CPU usages goes way up and ping times increase.
12:25<concha>leeching user? unlikely, i'm seeing the same trouble on two different linodes on different hosts.
12:26<jblack>Actually my incoming is up a bit over the last hour, but only to 10KB a sec.
12:27<rainkid>my latency is averaging 700+ms
12:28<npmr>it's route-specific
12:28<jblack>Thats probably mail that got delayed though.
12:28<rainkid>load average is low
12:28<afv-13>i'm getting over 2000 ms pings
12:30<npmr>afv-13, i'm pinging your ip
12:30<npmr>afv-13, it looks like it's your connection that's introducing the latency
12:30<afv-13>i've just started a download
12:30<anderiv>yah- my latency is around 35-40ms.
12:31<anderiv>12:06 -!- petert [] has joined #linode
12:31<jblack>looks like a dying router.
12:31<anderiv>nevermind that...
12:31<jblack>I'm seeing an average of 711ms, with a packet loss of 45%
12:31<npmr>jblack, between my linode and your ip, the problem is on the alternet link from dallas to new york
12:32<jblack>Yeah. At
12:32<npmr>roughly, yes
12:32<jblack>So linode only uses one backbone.
12:32<npmr>it uses several
12:32<jblack>That's uunet's.
12:33<npmr>and it looks like one of them is having an issue
12:34<jblack>bleh. I got tricked. It reverses to uunet and forwards to alternet's gateway.
12:34<npmr>alternet is uunet
12:35<anderiv>looks like it's better now, FWIW.
12:35<jblack>anderiv: Nah.
12:36<anderiv>yah - nevermind. It was better for 30 seconds or so but now it sucks again.
12:36<jblack>Luckily my daughter already finished most of her lit work for the day. I can put her on algebra.
12:47<@linbot>New news from forums: slowness in General Discussion <>
12:57<jblack>guess we're not the only ones to notice.
12:57|-|SpaceHobo [] has joined #linode
12:58<tag>I can't register in the forums...annoying
12:59|-|petert [] has quit [Quit: Connection reset by peter]
12:59<jblack>Gah. It needs a seperate account for the forums.
13:06<jblack>all better. :)
13:06|-|jblack [] has quit [Quit: leaving]
13:19<concha>things have been back to normal for me here in Boston for several minutes
13:31|-|Nigel_ [] has quit [Ping timeout: 480 seconds]
13:47<gpd><randomness> where do people look for reviews of laptops -- and what manufacturers are top of your list?
13:49|-|anderiv [] has quit [Server closed connection]
13:49|-|anderiv [] has joined #linode
13:55|-|javafueled [] has joined #linode
13:55|-|javafueled [] has left #linode []
14:04|-|linville [] has quit [Quit: Leaving]
14:12|-|linville [] has joined #linode
14:17|-|rainkid [] has quit [Quit: CGI:IRC]
14:41|-|avid [] has joined #linode
14:48<@linbot>New news from forums: Slackware 11.0 in Feature Request/Bug Report <>
14:53|-|concha [] has quit [Ping timeout: 480 seconds]
14:59|-|anderiv [] has quit [Quit: leaving]
15:16|-|Newsome [] has quit [Quit: Linux: Now with employee pricing!]
15:25|-|Dreamer3 [] has joined #linode
15:30|-|avid [] has quit [Read error: Connection reset by peer]
15:30<@mikegrb>gpd:, apple
15:30[~]mikegrb runs
15:30|-|avid [] has joined #linode
15:32<encode>mikegrb: ahahah
15:36<warewolf>linbot: rr
15:36<@linbot>warewolf: *click*
15:58|-|_andrew_j_w [] has quit [Quit: Konversation terminated!]
16:24<brc>host33 lagged
16:27|-|Netsplit <-> quits: kvandivo, soundsoldier, @linbot, encode, gpd, grummund, Xel
16:27|-|Netsplit over, joins: linbot
16:27|-|Netsplit over, joins: grummund
16:27|-|Netsplit over, joins: soundsoldier
16:27|-|Netsplit over, joins: gpd
16:27|-|Netsplit over, joins: kvandivo
16:27|-|Netsplit over, joins: Xel
16:27|-|Netsplit over, joins: encode
16:27|-|NetHack #linode [+o linbot] by []
16:28|-|Netsplit <-> quits: afv-13, heidi, Battousai, sec39, npmr, @caker
16:28|-|mode/#linode [-o linbot] by ChanServ
16:28|-|Netsplit over, joins: heidi
16:28|-|Netsplit over, joins: afv-13
16:28|-|Netsplit over, joins: sec39
16:28|-|Netsplit over, joins: Battousai
16:28|-|Netsplit over, joins: caker
16:28|-|Netsplit over, joins: npmr
16:28|-|NetHack #linode [+o caker] by []
16:40|-|afv-13 [] has quit [Remote host closed the connection]
16:47|-|avid [] has quit [Quit: avid]
17:31|-|javafueled [] has joined #linode
17:31|-|javafueled [] has left #linode []
17:42|-|spr [] has quit [Quit: Spoon!]
17:43|-|spr [] has joined #linode
17:50|-|tag [] has quit [Quit: leaving]
18:48|-|Kurt [] has joined #linode
19:16|-|SpaceHobo [] has quit [Ping timeout: 480 seconds]
19:57<Kurt>oh, kick ass
19:57<Kurt>I just got the gig as principal trombonist for the Evansville Opera and Ballet Company
19:57<Kurt>beat out twelve other guys
19:59<encode>nice work
20:12|-|Kurt [] has quit [Ping timeout: 480 seconds]
20:23|-|linville [] has quit [Quit: Leaving]
20:31|-|Newsome [] has joined #linode
20:55|-|Newsome [] has quit [Quit: Linux: Now with employee pricing!]
21:11|-|kokoko [] has joined #linode
21:17|-|konoko [] has quit [Ping timeout: 480 seconds]
21:18|-|Eman [] has quit [Ping timeout: 480 seconds]
21:19|-|Eman [] has joined #linode
21:43|-|avid [] has joined #linode
21:49<encode>they're expensive though :S
21:51<Dreamer3>worth it i think
21:51<Dreamer3>i'm selling my g5 tower
21:51<Dreamer3>slimming down
21:52<encode>my mac mini works ok for now
21:52<encode>my next mac will probably be a macbook of some description
21:53<Dreamer3>i just got one of those the other day
21:53<Dreamer3>very nice
21:53<Dreamer3>gotta sell my ibook now too :-)
21:54<encode>how do you have so much money
21:56<Dreamer3>i think
21:56<Dreamer3>that and i've gotten decent at Ruby on Rails and keep busy with contract work full-time
21:56<Dreamer3>though selling the old computers will recoupe a lot of the cost of the new ones :-)
22:06<soundsoldier>how do you like your apple cinema?
22:07|-|FireSlash` [] has joined #linode
22:08|-|FireSlash [] has quit [Ping timeout: 480 seconds]
22:25|-|Dreamr3 [] has joined #linode
22:29|-|avid [] has quit [Quit: avid]
22:30|-|Dreamer3 [] has quit [Ping timeout: 480 seconds]
22:30|-|Dreamr_3 [] has quit [Ping timeout: 480 seconds]
22:30|-|Dreamer3 [] has joined #linode
22:36|-|Netsplit <-> quits: Internat, tierra, AndyHat
22:37|-|Netsplit over, joins: tierra
22:59|-|VS_ChanLog [] has left #linode [Rotating Logs]
22:59|-|VS_ChanLog [] has joined #linode
23:12|-|JonR [] has joined #linode
23:12<JonR>I can't log into LISH on host17. something wrong?
23:15<@caker>JonR: one sec
23:16<taupehat>caker: when you get a chance, I've got a somewhat offtopic but very geeky question I thought you might be able to answer
23:16<@caker>taupehat: go for it
23:16<taupehat>IOS commands
23:16<taupehat>what the heck is the difference between copy run start and write mem?
23:18<@caker>I've alwayhs just used write, but looks like copy run start does the same thing to me..
23:18[~]taupehat is mangling a ton of 3500-series switches at work of late, and the subject has been bugging him
23:19<taupehat>I guess the one thing about "copy run" is that you can write the running config out to a different file
23:19<taupehat>but not just any file - it throws an "invalid filename" error if you try to make one up
23:19<taupehat>copy run foo
23:19<@caker>"Type write mem or copy running startup to commit the changes" <-- example, but recurring google exceprts
23:19<taupehat>cop ru s
23:20<taupehat>also works
23:20[~]JonR grumbles.
23:20<JonR>I'm not out of ram, cpu, or io tokens.
23:20<@caker>JonR: no luck stll?
23:20<JonR>but my linode is still crawling along
23:20<JonR>caker: nope!
23:20[~]caker digs further
23:20<JonR>no packet loss, either, as far as I can tell
23:20<taupehat>mayhap someone's thrashing his host
23:20<JonR>load average: 20.38, 20.78, 15.98
23:21<@caker>JonR: which username?
23:21<taupehat>JonR: what's got you so cpu-bound?
23:21<taupehat>load average: 0.15, 0.05, 0.02
23:21<JonR>Cpu(s): 2.4% user, 1.6% system, 0.0% nice, 96.1% idle
23:21<JonR>i'm getting hammered by trac-spammers.
23:21<@caker>JonR: you're burning io somewhere
23:21<taupehat>JonR: when I was having problems with similar, my solution was to halt apache for a while
23:22<taupehat>like five or ten minutes
23:22<JonR>my io_tokens seems nearly full.
23:22<taupehat>the spammers went away
23:22<JonR>taupehat: that's only a temporary solution.
23:22<JonR>they come back in a day or so
23:22<taupehat>yeah, you'll have to implement fancier antispam
23:22[~]taupehat would do unspeakable things to the spammers, and not things they'd enjoy
23:23<@caker># iocheck skip
23:23<@caker>OK Filename Type Size Used Priority
23:23<@caker>/dev/ubdb partition 132088 90032 -1
23:23<@caker>OK io_count=173049610 io_rate=1003 io_tokens=393463 token_refill=512 token_max=400000
23:23<JonR>[skip@host17 lish]# io_status
23:23<JonR>OK io_count=173055798 io_rate=1406 io_tokens=389323 token_refill=512 token_max=400000
23:23<JonR>I did manage to get into lish now
23:23<@caker>Your node is the only one running now
23:23<JonR>am I maybe filling the network IO?
23:23<JonR>I don't know how to get a meter on that...
23:23<taupehat>that seems unlikely
23:23[~]JonR shuts down apache
23:23<@caker>no, this is disk bound
23:24<@caker>plus, you're +90M into swap, which is bad on a 100M box
23:24[~]JonR grumbles.
23:24<taupehat>Swap: 96 0 96
23:25<taupehat>I've got 12M ram phys available, wonder if that's enough to turn SA on
23:25<taupehat>I'm getting several phish mails per day now
23:25<@caker>JonR: easy to find: ps auxhf | grep D (usualy that'll work)
23:25<@caker>also, excuse my typing mistakes -- still getting used to this laptop
23:25<taupehat>caker: what'd you get?
23:26<@caker>17" MacBook Pro (Core 2 Duo) !
23:26[~]taupehat notes that caker's ps example isn't much good when the date is "Dec02"
23:26<taupehat>nice laptop!
23:26<taupehat>ps auxhf | grep -e D$
23:26<@caker>bigger monitor for the win
23:27[~]efudd now sees a huge 960x600 "HELLO WORLD" on his sony sxrd from the xbox360 ><
23:27<efudd>TAKE THAT.
23:27<@caker>depending on ps version, you might need grep DW
23:27<efudd>WTb, system.outline() :/
23:27<efudd>SpriteBatch and Texture2D makes efudd sadface.
23:28<taupehat>I'm rather taken with the joys of POE switching
23:28<taupehat>we've got phones, access points, etc. all not needing a power cable run to them
23:28<taupehat>fewer cables FTW
23:28<taupehat>and so far, the poweron handshake has worked perfectly - haven't nuked anything
23:28<JonR>it's taking forever to shut down apache
23:29<efudd>ok, now i need a 1920x1080 BSOD image
23:29<efudd>to push off of the xbox360 as a "game"
23:29[~]efudd hoho
23:29<taupehat>JonR: killall -9 httpd
23:29<tierra>caker: I may not have a MacBook, but I like my 20" iMac 2Ghz Intel Core Duo at work ;)
23:29<taupehat>or apache2
23:30<@caker>tierra: good stuff
23:30[~]taupehat scowls at his latitude
23:30<@caker>the track pad trickery is, well .. trick
23:30<taupehat>heh, two fingers, man, two fingers =]
23:31<JonR>okay, what am I looking for in the 'ps auxhf | grep D' ?
23:31<taupehat>JonR: whatever's hogging ram
23:31<@caker>well, that *might* give you what's in D state, waiting on disk
23:32<JonR>trac, postgres, courier, awstats...
23:32[~]efudd has naked chick displayed on TV through xbox 360
23:32<efudd>take that!
23:32<efudd>I <3 this XNA
23:37<JonR>okay, it seems apache and two webapps are singlehandedly responsible for my linode being 90 megs into swap.
23:37<taupehat>JonR: I'll have to dig up my old post on what I did bad
23:37<@caker>JonR: pretty typical
23:37<@caker>JonR: time to futz with MaxClients and friends
23:38<JonR>or move trac to a local server.
23:38<JonR>because I'm the only person who uses it anyway.
23:38<JonR>nobody cares about my OSS projects. :D
23:38<taupehat>don't feel too bad =]
23:38<taupehat>hehe, caker who did the xmas ornaments on the logo?
23:39<@caker>TJ Frevert, friend o mine who helped make the initial logo
23:40<taupehat>right non
23:40<taupehat>is somewhat google of you =]
23:40<@caker>bah humbug
23:41<@caker>I'd prefer it to be a little more Holiday and a little less Christmas
23:41<taupehat>that whole mess
23:42<taupehat>I've gotten to the point where I say "Merry Xmas" to people, pronouncing the "X"
23:42<taupehat>a few people have asked, and I explain that you can't win - say Merry Christmas and people are offended, say Happy Holidays and other people are offended
23:42<taupehat>Xmas is sort of like taking "X" for a last name
23:43<tierra>I'm offended that you'd say "Merry Xmas"!
23:43<JonR>I have 40 mb free now, and 22 mb swap used.
23:43<JonR>I wonder why it doesn't just load all that swapped out memory into the free ram?
23:43<taupehat>JonR: you should read my link when you get a chance
23:44<JonR>I did.
23:44<@caker>JonR: and amazingly, the host load is back to normal!
23:44[~]JonR isn't using any java, thankfully.
23:44<JonR>caker: I'm removing the trac setup from my server.
23:44<taupehat>JonR: welcome to the node-thrashers club
23:44<JonR>if I have to, I'll remove the other webapp as well.
23:44<taupehat>you get to do the head-banger thing now
23:45<@caker>JonR: without knowing more about your setup, my guess now is that you'd be OK if you tweaked apache a bit (?)
23:45<JonR>caker: I've already applied the tweaks I found in the forum
23:46<JonR>MinSpareServers 5
23:46<JonR>MaxSpareServers 10
23:46<JonR>StartServers 5
23:46<JonR>MaxClients 50
23:46<JonR>MaxRequestsPerChild 32
23:46<JonR>I'm running apache 1.3
23:47<@caker>MaxClients is way too high, IMO
23:47<JonR>also KeepAliveTimeout 20 and MaxKeepAliveRequests 32
23:47<JonR>what should it be? 10?
23:47<@caker>I'd change MaxClients and MaxSpareServers down to 10 or less
23:47<@caker>the rest are OK, I think
23:53[~]taupehat has to comment on the stupidest thing post as he's done another
23:55|-|FireSlash` [] has quit [Read error: Connection reset by peer]
23:59<taupehat>linbot, hurry up and spam the link =]
---Logclosed Tue Dec 12 00:00:02 2006