Back to Home / #uml / 2007 / 04 / Prev Day | Next Day
#uml IRC Logs for 2007-04-18

---Logopened Wed Apr 18 00:00:15 2007
00:40|-|ElectricElf [] has quit [Ping timeout: 480 seconds]
00:45|-|ElectricElf [] has joined #uml
01:43|-|Ancalagon [] has joined #uml
02:26|-|m0sh [] has joined #uml
02:43|-|motp [~motp@] has joined #uml
03:18|-|aroscha [] has quit [Quit: aroscha]
03:47|-|aroscha [~aroscha@] has joined #uml
04:04|-|motp [~motp@] has quit [Quit: Leaving]
04:30|-|krau [~cktakahas@] has quit [Ping timeout: 480 seconds]
04:30|-|krau [~cktakahas@] has joined #uml
04:45|-|m0sh [] has quit [Ping timeout: 480 seconds]
05:47|-|aroscha [~aroscha@] has quit [Quit: aroscha]
05:53|-|aroscha [~aroscha@] has joined #uml
05:57|-|nikkne [] has joined #uml
06:42|-|polyonymous [] has quit [Ping timeout: 480 seconds]
07:13|-|polyonymous [] has joined #uml
07:28|-|nikkne [] has quit [Quit: nikkne]
07:49|-|nikkne [] has joined #uml
07:57|-|aroscha [~aroscha@] has quit [Quit: aroscha]
08:14|-|SNy [] has quit [Read error: Operation timed out]
08:36|-|aroscha [] has joined #uml
08:38|-|jdike [] has joined #uml
08:38<jdike>Hi guys
08:41<caker>good morning
08:45<aroscha>hi :)
08:45<aroscha>good afternoon
09:06|-|Ancalagon [] has quit [Quit: ChatZilla [Firefox]]
09:29[~]jdike adds per-device IO threads
09:34|-|nikkne [] has quit [Quit: nikkne]
09:44|-|baroni [] has quit [Remote host closed the connection]
09:46<dgraves>morning jdike.
09:46<dgraves>jdike: but do they deal well in fake smp?
09:46<jdike>you managed to swim to work
09:46<dgraves>i tried to find that island to rest in the merrimack.
09:46<dgraves>but i couldn't.
09:47<jdike>the locking is cleaner actually
09:47<jdike>since the devices are already locked
09:48<dgraves>behind bars!
09:48<dgraves>nice though.
09:48[~]dgraves is going to work for jdike when oracle gets tired of him.
09:50|-|baroni [] has joined #uml
09:51<jdike>now to add readv/writev support and see what that does
09:52|-|nikkne [] has joined #uml
09:59<dgraves>hopefully reads and writes?
09:59<dgraves>but you might get a bonus panic. :)
09:59<jdike>you better stay at Oracle
10:01|-|kokoko1 [~Slacker@] has joined #uml
10:05<dgraves>but i'm good at bonus panics!
10:07|-|baroni [] has quit [Remote host closed the connection]
10:39<jdike>this requires more thought
10:39<jdike>an iovec doesn't have an offset
10:48|-|kokoko1 [~Slacker@] has quit [Quit: Leaving]
11:28|-|nikkne [] has quit [Ping timeout: 480 seconds]
11:43|-|richardw [] has joined #uml
12:11|-|richardw_ [] has joined #uml
12:16|-|richardw [] has quit [Read error: Operation timed out]
12:18<dgraves>jdike: add one! :)
12:23<jdike>so full of useful suggestions
12:24<dgraves>i didn't realize this was my hiring test. :)
12:29|-|hfb [] has joined #uml
12:29|-|hfb [] has left #uml []
12:35|-|nikkne [] has joined #uml
12:35|-|nikkne [] has quit []
12:55<aroscha>jdike: is using a linux bridge (brctl etc..) an alternative to using an uml_switch (sorry if this question might sound stupid)?
12:56<jdike>all interfaces will see all packets, I think
12:56<aroscha>like I imagine that i have a bunch of tapX devices and those that I want to be connected I will simple brctl addbr to some brX that I created on the host
12:58<aroscha>so what you are saying is that it will be more like a hub then ?
12:59|-|richardw_ [] has quit [Quit: Leaving]
13:00<aroscha>BTW: I increased the socket backlog and /proc/sys/net/core/wmem_max and rmem_max to really high values... no success. I still get the fd 4 : Resource not available bug in uml_switch
13:07<jdike>just give each port its own socket
13:31|-|besonen_mobile_ [] has quit [Ping timeout: 480 seconds]
13:41|-|kokoko1 [~Slacker@] has joined #uml
13:59|-|dgraves [] has quit [Remote host closed the connection]
14:37|-|baroni [] has joined #uml
15:42<aroscha>how do you intentionally panic (and thus reboot) a linux system? Is there a syscall for it?
15:43<caker>panic or cause a reboot?
15:43<aroscha>like panic(int flags) or so ?
15:43<caker>echo b > /proc/sysrq-trigger
15:43<caker>oh .. hmm
15:43<aroscha>seems to work
15:43<caker>yeah, that does a force reboot
15:44<aroscha>thx! I will read that link
15:44<aroscha>now i dont have to travel to the box
15:44<caker>very useful in those situations
15:44<caker>many a time sysrq has saved my ass a trip (or worse: dealing with datacenter tech support)
15:45<aroscha>hehe, actually (i am making myself probably a person non grata here but...) the UML kernel sort of made every process a zombie
15:46<aroscha>including the reboot cmd
15:46<aroscha>so that's why i wanted it to panic (instead of me)
15:46<caker>was this one of your 9million uml instances on that machine? :)
15:47<aroscha>no, this was from the default UML kernel from the VNUML site
15:48<aroscha>well, i was curious and playing around . I _guess_ the VNUML kernel is a 32 bit. And well... yes, any process that I would start would do its work (or not) but at the end it would become a zombie . init was using 100% CPU
15:48<aroscha>I guess I have a lucky hand for crashing things
15:49<aroscha>hehe, I should hire at intel as a UML tester
15:52<jdike>aroscha, what version is that?
15:52<jdike>I don't remember ever seeing anything like that
15:53<aroscha>jdike: i simply downloaded linux- from the VNUML site
15:53<jdike>hmmm, OK
15:54<jdike>try a newer UML though
15:54<aroscha>unpacked it and started it in my host
15:54<aroscha>as I said, I was just trying things
15:54<aroscha>comparing my current network simulator to VNUML
15:56<jdike>np, just I'd prefer you apply your talent for breaking things to something I'm likely to fix
16:59|-|jdike [] has quit [Quit: Leaving]
17:12|-|kokoko1 [~Slacker@] has quit [Quit: Leaving]
18:35|-|turnp [] has quit [Remote host closed the connection]
20:59|-|aroscha [] has quit [Quit: aroscha]
22:52|-|besonen_mobile [] has joined #uml
22:59|-|VS_ChanLog [] has left #uml [Rotating Logs]
22:59|-|VS_ChanLog [] has joined #uml
---Logclosed Thu Apr 19 00:00:09 2007