| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Comments:
<0> wish I could look at the machine directly
<1> you can come over and look over my shoulder
<0> heh
<0> unless you're in Seattle, that probably won't work :P
<1> yeah no. ATL
<1> even if I shutdown the webserver on the box the load will stay at 12
<2> tried disabling ht?
<3> what app server are you using tomcat, weblogic, websphere?
<1> I'm trying to remember if we've done that or not
<0> that's just stabbing in the dark
<1> pramod: SunONE webserver 6.1sp3
<4> 3?
<4> There's a 3?
<1> there is a 6
<1> sp6 is the latest
<4> There's a service pack 6 for 6.1? WTF, I *just* downloaded it... unless I'm confused, I thought it said it was 2...
<0> I think you're being lied to
<1> blueroo: me or CeruleanD?
<0> you
<1> blueroo: uptime is lying to me?
<0> the numbers aren't adding up
<1> blueroo: well DUH! .. :)
<0> yes, but figuring out where is the trick :)
<4> No wait... it says on the site it's up to sp5.
<4> So I am confused, I think it was 2 I downloaded a month or two ago.
<1> CeruleanD: there are docs for SP6 and a release note
<4> Oh wait yeah I see 6 now, duh
<0> what's iostat say?
<3> if your load is staying at 12 even though you shutdown the webserver can you do a top and see what process is maxing out the cpu
<1> blueroo: http://pastebin.ca/81765 without webserver running
<1> and load according to uptime stays at 12
<3> can you do a top
<1> pramod: that's just it there isn't one. since CPU is idle
<1> pramod: hence that vmstat sez so
<0> context switching dropped to the floor, which is expected
<0> you're still doing a lot of interrupts for being so idle. handling packets most likely
<1> blueroo: yeah. probably nfs stuff
<0> you are NFS mounted?
<1> three to be exact
<0> can you get an iostat with the server off and on?
<1> any params you want?
<0> just general for now
<1> 5 5 or something else?
<0> how about every 1 second up to 10
<1> http://pastebin.ca/81773
<1> with webserver running
<0> are you logging to NFS or local disk?
<1> local
<0> pulling shared data from NFS for the app then
<1> yeah
<1> and even then the webserver will cache it in it's own filecache
<1> http://pastebin.ca/81775
<1> wihout webserver running
<0> of course
<0> iostat won't give us NFS stats
<0> that's teh ****
<1> yup
<0> do you have nfsstat?
<0> unfortunately I don't have a reference NFS capable machine handy
<1> yup
<0> well, you could always 'nfsstat -o all'
<0> better too much information than none at all
<1> didn't like that all facility
<0> just use -v then
<0> what version of 2.4 is this machine running, btw?
<1> oh this is 2.6.5
<1> 2.6.5-7.244-bigsmp SuSE 9.3
<1> http://pastebin.ca/81785
<1> that doesn't seem all that helpful
<0> you're right. it's nearly useless
<1> false boolean I guess.
<0> :)
<1> yeah and watching the numbers they aren't incrementing at an astonishing rate
<0> hot many threads are you running?
<1> well...
<1> just looking right now. 4 active threads handingling requests in the webserver
<1> but then there are another 20 request threads pulling requests off of the sockets and handing them to the worker threads
<1> then another 10 or 20 threads handeling subsustems in the webserver.
<1> then of course all the Java stuff running.. garbage collector and threads for the worker threads in the webserver that process the requests int eh web app
<1> that is rtfn at a low traffic time
<1> only handling around 4500 hpm on that one box
<1> she peaked out today around 9k hpm or so
<0> when you drop the webserver, the JVM goes down completely as well right?
<1> correct
<0> it's times like this where you wish for dtrace :)
<0> just out of curiosity, what does mpstat -P ALL say?
<1> http://pastebin.ca/81796
<1> yeah, dtrace would be nice
<0> this machine doesn't have any siblings, does it
<1> yeah. 16 machines total doing the same service
<0> and only this machine has the problem
<1> no. all 16 do
<1> they all exhibit the same thing.
<1> I'll need to hit the operations dudes up next week to see if any newer or updated versions of the SuSE stuff do the same thing.
<1> I don't know what we have installed where.
<0> hrm
<0> I'd be curious to see if they're all running the same kernel version
<0> whether the problem persists with the latest available SuSE kernel and JVM
<1> yes. all the same kernel version
<1> CeruleanD: and you know SunONE 7.0 is out for the public too I think
<4> Yeah, I saw that, but I needed it for a pair of production machines and my boss has qualms about using what's basically beta software for production. :)
<4> "Technology Preview" = fancy words for "beta"
<1> CeruleanD: yeah. I played with it a little, but haven't had a chance to really go through it
<5> SunONE is also renamed. Sun Java Webserver. From what I've seen so far (running onmy blade) its purrty.
<4> Yeah it is. Annoying.
<5> Better Java/JSP support and the administration interface has been merged with the application/access server. Its pretty decent IMO.
<4> How do you define a webserver as purty? :-P
<6> Hi. This is not working in Mozilla web page editor when I am trrying to publish the web site ftp://ftp.192.168.0.3/
<4> hmmmm
<5> frodo_: thats pretty obvious.
<5> CeruleanD: the way it looks, feels and it build. For example; the way you can use both the commandline and admin console, stuff like that.
<6> That is what mozilla is telling me to use.
<5> frodo_: I doubt it.
<4> You sure you don't want ftp://192.168.0.3 ?
<4> Huh, I'll have to play with it some more.
<6> I am looking the screen and that is what is saids.
<5> frodo_: Then its time to file a bugreport. Also using your brains would help.
<3> frodo_: does the ftp site have a username and p***word
<5> CeruleanD: well, its just a snapshot. So far I really like the previous release (SunONE 6.1SP5) better though.
<6> There is a area for that.
<6> There is nother fields for that.
<6> This is not working either ftp://192.168.0.3
<3> frodo: try ftp://username:p***word@192.168.0.3 where you replace username and p***word with your ftp username and p***word
<5> lol
<5> frodo_: so is there actually an ftp server running on that host?
<5> frodo_: do yourself a favor and RTFM.
<6> Thanks
<5> ofcourse it would help if you'd define "doesn't work" with more info (an error message would be nice) but I guess thats asking too much effort.
<4> frodo: Wait, I thought you said it's not working... yet it's asking for username/p***word?
<6> ftp://webmaster:my***word@192.168.0.3 is not working.
<5> in the "publish" screen it even has frickin' examples listed.
<5> "for example ftp://www.myisp.com". So he's full of bull****.
<6> Before you kick be right because the author of the program should use the correct examples on the screens.
<5> besides; this isn't Mozilla support either, for all we know the other host doesn't have a FTP server running but this idiot sure isn't going to tell us.
<3> frodo_: from a terminal can you try telnet 192.168.0.3 21 to see if the ftp server is running
<6> Your the ****ing retard
<5> frodo_: sure, sure.
<5> frodo_: the program HAS working examples you clueless twit.
<6> I have been writting programs for 15 years and they all work with no problems
<5> rofl
<5> I also fail to see the relevance there but hey; what do we know.
<7> I am having a problem trying to get X to run, i feel its a problem with my graphics settings.. in my xorg.conf under the 'graphics device settings' these are the settings, do u think these are correct for a generic setup
<7> Section "Device"
<7> Identifier "VESA Framebuffer"
<7> Driver "vesa"
<7> #VideoRam 4096
<7> # Insert Clocks lines here if appropriate
<3> bigdog: try Xorg -configure as root and X will create a sample xorg.conf that will will work for your configuration
<5> bigdog: don't ever do that again.
Return to
#linux
or
Go to some related
logs:
bigilla
#java
#chatzone
roxanne duncan
lupen cartoons
#AllNiteCafe
msdn clear is not a member of basic_string
#apache
#chatzone
mbmon em64
|
|