Windows LambdaLoadTest hang at 3% #8214
Comments
|
https://ci.eclipse.org/openj9/job/Test_openjdk8_j9_special.system_x86-32_windows_Personal/45 |
|
https://ci.eclipse.org/openj9/job/Test_openjdk8_j9_special.system_x86-32_windows_Personal/69 |
|
https://ci.eclipse.org/openj9/job/Test_openjdk8_j9_special.system_x86-32_windows_Personal/93 |
|
https://ci.eclipse.org/openj9/job/Test_openjdk11_j9_sanity.system_x86-64_windows_Nightly/320 |
pshipton
changed the title
|
https://ci.eclipse.org/openj9/job/Test_openjdk8_j9_special.system_x86-32_windows_Personal/99 |
|
https://ci.eclipse.org/openj9/job/Test_openjdk8_j9_special.system_x86-32_windows_Release/9 |
|
https://ci.eclipse.org/openj9/job/Test_openjdk8_j9_special.system_x86-32_windows_Personal/117 |
|
https://ci.eclipse.org/openj9/job/Test_openjdk8_j9_special.system_x86-32_windows_Release_lambdaLoadTest/1 https://ci.eclipse.org/openj9/job/Test_openjdk8_j9_special.system_x86-32_windows_Personal_lambdaLoadTest/5 |
|
https://ci.eclipse.org/openj9/job/Test_openjdk8_j9_special.system_x86-32_windows_Release_lambdaLoadTest/2 |
|
https://ci.eclipse.org/openj9/job/Test_openjdk8_j9_special.system_x86-32_windows_Personal_lambdaLoadTest/20 |
|
https://ci.eclipse.org/openj9/job/Test_openjdk8_j9_special.system_x86-32_windows_Personal_lambdaLoadTest/22 |
|
https://ci.eclipse.org/openj9/job/Test_openjdk8_j9_special.system_x86-32_windows_Personal_lambdaLoadTest/23 |
|
@Mesbah-Alam we've been ignoring this for while, but it would be worth taking a look. |
|
Created STF issue for the above: adoptium/STF#86 |
|
@pshipton FYI : Simon (@lumpfish) has created a fix (adoptium/STF#89) which should resolve the procdump hang issue. We may close this issue at this point and re-open if the problem somehow resurfaces in any future build. |
|
https://ci.eclipse.org/openj9/job/Test_openjdk8_j9_special.system_x86-32_windows_Personal_testList_1/1/ |
|
There are 3 core files, which I'll refer to as 125, 155, 225. 125: 0x139c: 0x33d8: 225: same threads 0x139c: 0x33d8: Running !stack on jdmpview results in: |
|
It seems it could be a decompilation endless loop. @0xdaryl fyi. Somebody could grab the diagnostics before they expire and get deleted. This problem has been occurring for a while, we've only "recently" been able to get core files. |
|
fwiw, I'm copying the sdk, debug image, and core files to /team/triage/openj9-8214. |
|
Also copied them to Box, pls let me know if you need access. |
|
@nbhuiyan : could you take an initial look at this please? I realize you're about to vanish this week on vacation, but perhaps a quick triage from the crash dump file could be done. Thanks. |
|
@0xdaryl Looking at the core dumps briefly, I also suspect the hang is happening during decompilation. However, further work is needed to determine the cause and coming up with a fix. Requesting that this issue is moved to the jdk16 release. I have kept all the files needed to continue this investigation upon my return. |
|
Moving forward for the justification above. |
0xdaryl
modified the milestones:
Release 0.24 (Java 8, 11, 15) Jan refresh,
Release 0.25 (Java 16)
|
https://ci.eclipse.org/openj9/job/Test_openjdk8_j9_special.system_x86-32_windows_Personal_testList_0/6 |
|
https://ci.eclipse.org/openj9/job/Test_openjdk8_j9_special.system_x86-32_windows_Personal_testList_0/7/ |
|
https://ci.eclipse.org/openj9/job/Test_openjdk8_j9_special.system_x86-32_windows_Personal_testList_0/8/ |
pshipton
changed the title
|
@0xdaryl, I have determined a possible cause for this hang from what I could gather from the diagnostics files over the last month. I am still working on the fix and will need more time to test and verify the issue has been resolved, which will take time as this hang happens intermittently. Requesting that this issue is moved forward to the next milestone. |
pshipton
modified the milestones:
Release 0.25 (Java 16),
Release 0.26 (Java 8, 11, 16) Apr refresh
|
Moved it. This is not a new issue, I'm glad we're getting to the bottom of it. |
|
@nbhuiyan is working on a fix and studying its behaviour in local testing before contributing. If it is committed, he requested more time to allow it to "soak" in the head stream. I will push this out to 0.27. |
0xdaryl
modified the milestones:
Release 0.26 (Java 8, 11, 16) Apr refresh,
Release 0.27 (Java 8, 11, 16) July refresh
|
I haven't seen this in some time, did a fix get merged? |
|
I have not been able to verify that the possible fix I had would address this hang, since over the last month in my local testing using a recent build, I did not experience this hang without the fix. It is possible that this got fixed as a side effect of some other work. I suggest we close this issue now. If we see this hang again, we can revisit this issue. |
https://ci.eclipse.org/openj9/job/Test_openjdk8_j9_special.system_x86-32_windows_Personal/33
LambdaLoadTest_OpenJ9_NonLinux_special_24
variation: Mode107-OSRG
JVM_OPTIONS: -Xgcpolicy:optthruput -Xdebug -Xrunjdwp:transport=dt_socket,address=8888,server=y,onthrow=no.pkg.foo,launch=echo -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation
No diagnostic files generated.
The text was updated successfully, but these errors were encountered: