No SIGINT with 1.4.2-rc1?
Hi, I've just upgraded from 1.3.0 to Blackdown-1.4.2-rc1 and am having some grief. I've tried googling on this topic with no success.. Since I upgraded the JVM, some applications can no longer be terminated with control-c! Control-\ no longer dumps information either. Also, when I manually kill the java processes, only the particular pids that I specify are killed. With the previous JVM I could kill one of the parent java processes and the whole group would be terminated. The fault might be triggered by particular run-time or start-up conditions. The same application that was unresponsive to control-c earlier is now responding to it (after a restart, ie different program instance but same program code). The applications that have been affected are all heavily threaded FWIW. I've read that the JVM has a -Xrs option. I am NOT using that. I've seen the same behavior on at least two different computers. Both machines are Debian stable/testing. I downloaded the JVM which was built with gcc-3.2. I'd love to hear any suggestions. It is very hard to restart these applications ATM! Many thanks, Dave -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: No SIGINT with 1.4.2-rc1?
Just for the record: This problem has not happened again for a few months. I think it was associated with the libstdc++ upgrade that accompanied my Java upgrade. Sorry for my confusion. Cheers, Dave On Mon, 2004-09-13 at 14:08, David Brodrick wrote: > Hi, > I've just upgraded from 1.3.0 to Blackdown-1.4.2-rc1 and am having some > grief. I've tried googling on this topic with no success.. > > Since I upgraded the JVM, some applications can no longer be terminated > with control-c! Control-\ no longer dumps information either. Also, when > I manually kill the java processes, only the particular pids that I > specify are killed. With the previous JVM I could kill one of the parent > java processes and the whole group would be terminated. > > The fault might be triggered by particular run-time or start-up > conditions. The same application that was unresponsive to control-c > earlier is now responding to it (after a restart, ie different program > instance but same program code). The applications that have been > affected are all heavily threaded FWIW. > > I've read that the JVM has a -Xrs option. I am NOT using that. > > I've seen the same behavior on at least two different computers. Both > machines are Debian stable/testing. I downloaded the JVM which was built > with gcc-3.2. > > I'd love to hear any suggestions. It is very hard to restart these > applications ATM! > > Many thanks, > Dave > > > > -- > To UNSUBSCRIBE, email to [EMAIL PROTECTED] > with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Any resources shared between JVMs?
Hi All, I'm chasing a problem where a (normally) smoothly-running java program seems to briefly stall from time-to-time. The CPU does not appear exhausted (according to "top") but the machine is still fairly busy. One of the things that makes the machine busy is a program that fires up a different java program in a tight loop. I was wondering whether different JVMs on the same host can interact in some way? Are there any common resources that could enable one JVM to be blocked for a short period while another JVM is starting up? Or is my problem more likely to be normal resource (CPU) competition? I suspect it's the latter but thought I'd ask the experts to be sure. I'm running: Java(TM) 2 Runtime Environment, Standard Edition (build Blackdown-1.4.2-rc1) Java HotSpot(TM) Client VM (build Blackdown-1.4.2-rc1, mixed mode) Many thanks, Dave -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Any resources shared between JVMs?
My issue turned out to be a simple matter of CPU exhaustion after all. There's something funny with the threads on 2.6 kernels and procps (at least the one in Debian Sarge) in that top will only show the parent process and not show any of the threads. Therefore the CPU was actually much busier than one would think by looking at the process list in top. Thanks very much for your help though. Dave On Thu, 2006-10-12 at 09:04 +1000, David Brodrick wrote: > Hi All, > I'm chasing a problem where a (normally) smoothly-running java program > seems to briefly stall from time-to-time. > > The CPU does not appear exhausted (according to "top") but the machine > is still fairly busy. One of the things that makes the machine busy is a > program that fires up a different java program in a tight loop. > > I was wondering whether different JVMs on the same host can interact in > some way? Are there any common resources that could enable one JVM to be > blocked for a short period while another JVM is starting up? Or is my > problem more likely to be normal resource (CPU) competition? > > I suspect it's the latter but thought I'd ask the experts to be sure. > I'm running: > Java(TM) 2 Runtime Environment, Standard Edition (build > Blackdown-1.4.2-rc1) > Java HotSpot(TM) Client VM (build Blackdown-1.4.2-rc1, mixed mode) > > Many thanks, > Dave > > > -- > To UNSUBSCRIBE, email to [EMAIL PROTECTED] > with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
