fiorbd test failed

[liaow8]

when I start benchmark test,faild,how to resolve this problem [2017-10-17T08:33:28.541710][LOG]: start to run performance test [2017-10-17T08:33:28.546344][LOG]: Calculate Difference between Current Ceph Cluster Configuration with tuning [2017-10-17T08:33:33.438428][LOG]: Tuning[analyzer] is not same with current configuration [2017-10-17T08:33:33.908256][LOG]: Tuning has applied to ceph cluster, ceph is Healthy now [2017-10-17T08:33:36.914239][LOG]: ============start deploy============ [2017-10-17T08:33:39.591453][LOG]: Shutting down mon daemon [2017-10-17T08:33:40.237892][LOG]: Shutting down osd daemon [2017-10-17T08:33:40.576835][LOG]: Starting mon daemon [2017-10-17T08:33:40.943608][LOG]: Started mon.node02 daemon on node02 [2017-10-17T08:33:41.319379][LOG]: Started mon.node03 daemon on node03 [2017-10-17T08:33:41.696928][LOG]: Started mon.node01 daemon on node01 [2017-10-17T08:33:41.697034][LOG]: Starting osd daemon [2017-10-17T08:33:42.005184][LOG]: Started osd.0 daemon on node01 [2017-10-17T08:33:42.317962][LOG]: Started osd.1 daemon on node01 [2017-10-17T08:33:42.634781][LOG]: Started osd.2 daemon on node02 [2017-10-17T08:33:42.954045][LOG]: Started osd.3 daemon on node03 [2017-10-17T08:33:43.432979][LOG]: not need create mgr [2017-10-17T08:33:43.443442][LOG]: Clean process log file. [2017-10-17T08:33:43.919403][WARNING]: Applied tuning, waiting ceph to be healthy [2017-10-17T08:33:47.403901][WARNING]: Applied tuning, waiting ceph to be healthy [2017-10-17T08:33:50.888564][LOG]: Tuning has applied to ceph cluster, ceph is Healthy now [2017-10-17T08:33:52.350882][LOG]: RUNID: 13, RESULT_DIR: //mnt/data//13-3-fiorbd-seqwrite-4k-qd64-2g-100-400-rbd [2017-10-17T08:33:52.351263][LOG]: Prerun_check: check if sysstat installed [2017-10-17T08:33:52.658104][LOG]: Prerun_check: check if blktrace installed [2017-10-17T08:33:53.332501][LOG]: check if FIO rbd engine installed [2017-10-17T08:33:53.720802][LOG]: check if rbd volume fully initialized [2017-10-17T08:33:54.206562][WARNING]: Ceph cluster used data occupied: 2.698 KB, planned_space: 10485760.0 KB [2017-10-17T08:33:54.206722][WARNING]: rbd volume initialization has not be done [2017-10-17T08:33:54.206871][LOG]: Preparing rbd volume [2017-10-17T08:33:55.164264][LOG]: 1 FIO Jobs starts on node02 [2017-10-17T08:33:55.474774][LOG]: 1 FIO Jobs starts on node03 [2017-10-17T08:33:55.783428][LOG]: 1 FIO Jobs starts on node01 [2017-10-17T08:33:57.122272][WARNING]: 0 fio job still runing [2017-10-17T08:33:57.122398][ERROR]: Planed to run 0 Fio Job, please check all.conf [2017-10-17T08:33:57.123074][ERROR]: The test has been stopped, error_log: Traceback (most recent call last): File "/CeTune/benchmarking/mod/benchmark.py", line 46, in go self.prerun_check() File "/CeTune/benchmarking/mod/bblock/fiorbd.py", line 89, in prerun_check self.prepare_images() File "/CeTune/benchmarking/mod/bblock/fiorbd.py", line 52, in prepare_images raise KeyboardInterrupt KeyboardInterrupt

[xuechendi]

@liaow8 , hi, thanks for your report, looks like there are some incorrect configuration, which led to fio unable to run test, can you send logs under CeTune/Log/ to me? Or if you're familiar with python and fio, you can check your self, the process_log shows all command being executed during test, so it is easy to find out what is the really failure point.

[liaow8]

sorrt,I'm not familiar with python an fio,I send log for you cetune_python_log_file.log cetune_error_log_file.log cetune_process_log_file.log

[xuechendi]

Hi,

I noticed that cetune gracefully interrupted that failure, so logs are moved to //mnt/data//15-3-fiorbd-seqwrite-4k-qd64-2g-100-400-rbd, can you also send out logs under that folder?

Best regards, Chendi

From: wei liao [mailto:notifications@github.com] Sent: Tuesday, October 17, 2017 12:46 PM To: 01org/CeTune CeTune@noreply.github.com Cc: Xue, Chendi chendi.xue@intel.com; Comment comment@noreply.github.com Subject: Re: [01org/CeTune] fiorbd test failed (#274)

sorrt,I'm not familiar with python an fio,I send log for you cetune_python_log_file.loghttps://github.com/01org/CeTune/files/1389767/cetune_python_log_file.log cetune_error_log_file.loghttps://github.com/01org/CeTune/files/1389768/cetune_error_log_file.log cetune_process_log_file.loghttps://github.com/01org/CeTune/files/1389769/cetune_process_log_file.log

— You are receiving this because you commented. Reply to this email directly, view it on GitHubhttps://github.com/01org/CeTune/issues/274#issuecomment-337116564, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AEJ1pqu6KRVbY_eRqExxkf4Shl3mbr3Tks5stDEfgaJpZM4P7bpb.

[liaow8]

please see the new logs,thank you cetune_python_error_log_file.log cetune_python_log_file.log cetune_process_log_file.log

[xuechendi]

@liaow8 , hi, through log, I saw your test is running normally, from 2017/10/17 09:47 - 09:55 without error interrupts, can you check ?

[liaow8]

yes,I found the problem.But I don't know the reason.I have 3 osd nodes,when I set the list_client to be one of the nodes in cluster configuration,the test can be success ,when I set the list_client to be 3 nodes,the test go to failed.I don't know the reason and how to set the list_client

[xuechendi]

Make sure all 3 nodes need to be able to auto ssh by head, and all three nodes install fio, ceph rbd.

And if you can reproduce the issue, please send me those logs.

From: wei liao [mailto:notifications@github.com] Sent: Tuesday, October 17, 2017 2:29 PM To: 01org/CeTune CeTune@noreply.github.com Cc: Xue, Chendi chendi.xue@intel.com; Comment comment@noreply.github.com Subject: Re: [01org/CeTune] fiorbd test failed (#274)

yes,I found the problem.But I don't know the reason.I have 3 osd nodes,when I set the list_client to be one of the nodes in cluster configuration,the test can be success ,when I set the list_client to be 3 nodes,the test go to failed.I don't know the reason and how to set the list_client

— You are receiving this because you commented. Reply to this email directly, view it on GitHubhttps://github.com/01org/CeTune/issues/274#issuecomment-337131157, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AEJ1pkEZA17FipRlyAoKgNAQCrCBvWuuks5stEkdgaJpZM4P7bpb.

[liaow8]

I reproduce the problem,please see the logs cetune_python_log_file.log cetune_process_log_file.log cetune_error_log_file.log