Bug 43654 - java process spinning and using 100% of a cpu after test completes
Summary: java process spinning and using 100% of a cpu after test completes
Status: RESOLVED WORKSFORME
Alias: None
Product: JMeter - Now in Github
Classification: Unclassified
Component: Main (show other bugs)
Version: 2.3
Hardware: PC other
: P2 normal with 1 vote (vote)
Target Milestone: ---
Assignee: JMeter issues mailing list
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2007-10-18 11:00 UTC by RH
Modified: 2008-11-08 14:45 UTC (History)
0 users



Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description RH 2007-10-18 11:00:30 UTC
Test/server specifics follow after strace output.

Here's a snippet from the java process using strace on linux:
(I've seen similar spinning on win32 also)

[pid   396] gettimeofday({1192565615, 306474}, NULL) =
0
[pid   305] <... futex resumed> )       = 1
[pid   321] <... futex resumed> )       = 0
[pid   396] futex(0x80564a0, FUTEX_WAIT, 2, NULL
<unfinished ...>
[pid   305] gettimeofday( <unfinished ...>
[pid   321] futex(0x80564a0, FUTEX_WAKE, 1 <unfinished
...>
[pid   396] <... futex resumed> )       = -1 EAGAIN
(Resource temporarily unavailable)
[pid   396] futex(0x80564a0, FUTEX_WAKE, 1) = 0
[pid   305] <... gettimeofday resumed> {1192565615,
306909}, NULL) = 0
[pid   321] <... futex resumed> )       = 0
[pid   396] futex(0x80575d0, FUTEX_WAIT, 2, NULL
<unfinished ...>
[pid   305] futex(0x80575d0, FUTEX_WAIT, 2, NULL
<unfinished ...>
[pid   321] futex(0x80575d0, FUTEX_WAKE, 1 <unfinished
...>
[pid   305] <... futex resumed> )       = -1 EAGAIN
(Resource temporarily unavailable)
[pid   321] <... futex resumed> )       = 1
[pid   318] <... futex resumed> )       = 0
[pid   305] futex(0x80575d0, FUTEX_WAKE, 1 <unfinished
...>
[pid   321] gettimeofday({1192565615, 309070}, NULL) =
0
[pid   321] futex(0x80575d0, FUTEX_WAIT, 2, NULL
<unfinished ...>
[pid   322] <... futex resumed> )       = 0
[pid   305] <... futex resumed> )       = 1
[pid   322] futex(0x80575d0, FUTEX_WAIT, 2, NULL
<unfinished ...>
[pid   305] gettimeofday({1192565615, 309863}, NULL) =
0
[pid   305] futex(0x80575d0, FUTEX_WAIT, 2, NULL
<unfinished ...>
[pid 32765] <... futex resumed> )       = -1 ETIMEDOUT
(Connection timed out)
[pid 32765] futex(0x67d5e128, FUTEX_WAKE, 1) = 0




------------------------------
Testing and server specifics:
------------------------------
Jmeter 2.3

bin/ApacheJMeter.jar -n -H server1 -P 80 -t /jmeter/tst.jmx -l /jmeter/tst.jtl
-Jjmeter.save.saveservice.data_type=true -Jjmeter.save.saveservice.label=true
-Jjmeter.save.saveservice.response_code=true -Jjmeter.save.save
service.response_data=true -Jjmeter.save.save service.response_message=true
-Jjmeter.save.saveservice.successful=true
-Jjmeter.save.saveservice.thread_name=true -Jjmeter.save.saveservice.time=true
-Jjmeter.save.saveservice.timestamp_format=yyyy-MM-dd HH:mm:ss

jvm_args: -Xms756m -Xmx1224m -verbose:gc -XX:+PrintTenuringDistribution
-Xloggc:gc.log
OS:Red Hat Enterprise Linux AS release 4 (Nahant Update 5)
uname:Linux 2.6.9-55.0.6.ELhugemem #1 SMP Thu Aug 23 11:22:53 EDT 2007 i686
libc:glibc 2.3.4 NPTL 2.3.4
rlimit: STACK 10240k, CORE 0k, NPROC 274432, NOFILE 1024, AS infinity
load average:1.30 0.68 0.36
CPU:total 8 (4 cores per cpu, 1 threads per core) family 6 model 15 stepping 7,
cmov, cx8, fxsr, mmx, sse, sse2, sse3, ssse3
Memory: 4k page, physical 16631028k(13404036k free), swap 8388576k(8388576k free)
vm_info: Java HotSpot(TM) Server VM (1.6.0_03-b05) for linux-x86, built on Sep
24 2007 22:32:39 by "java_re" with gcc 3.2.1-7a (J2SE release)
Comment 1 Sebb 2007-10-22 03:00:59 UTC
Thans for the info so far, but there's not enough to draw any conclusions.

Can you attach a copy of the JMX file? If it contains any private information, 
be sure to change that first.

Does it happen with different tests, or only one?
Does it happen every time?
How many threads?
Any message in jmeter.log when ii does happen?
Are there any threads still running? - this can be seen from the closedown 
messages. There should be two messages for each, for example:

jmeter.threads.JMeterThread: Thread Thread Group 1-1 is done
jmeter.engine.StandardJMeterEngine: Ending thread Thread Group 1-1 
Comment 2 Sebb 2008-11-08 14:45:44 UTC
This has not been reported by anyone else, and it does not seem to be reproducible.
Comment 3 The ASF infrastructure team 2022-09-24 20:37:40 UTC
This issue has been migrated to GitHub: https://github.com/apache/jmeter/issues/2037