diff options
author | Yujun Zhang <zhang.yujunz@zte.com.cn> | 2016-12-31 21:27:38 +0800 |
---|---|---|
committer | Yujun Zhang <zhang.yujunz@zte.com.cn> | 2017-01-11 10:23:14 +0800 |
commit | 23f48e46a46976ae6f6d97aea11440e6a8b63121 (patch) | |
tree | 482573f3d64e1e7fcb3d054d109033d6a1b6fa82 /opt/plugin/doctor/collector.yaml | |
parent | 9fb9d1357de8c6f489b17b52c5919e978c3c930a (diff) |
sample QPI for doctor performance measured in verification job
The doctor benchmark plan produces a performance profile consist of
timestamp at each checkpoint and time cost of each phase.[1]
It collects data from log files and render a timeline diagram in console
Current progress
- skeleton created
- method to be implemented
[1]: http://artifacts.opnfv.org/doctor/docs/design/index.html#document-performance-profiler
JIRA: QTIP-199
Change-Id: I60fbbd37d04a5724841bd85cb3bfed9eabd3741a
Signed-off-by: Yujun Zhang <zhang.yujunz@zte.com.cn>
Diffstat (limited to 'opt/plugin/doctor/collector.yaml')
-rw-r--r-- | opt/plugin/doctor/collector.yaml | 24 |
1 files changed, 24 insertions, 0 deletions
diff --git a/opt/plugin/doctor/collector.yaml b/opt/plugin/doctor/collector.yaml new file mode 100644 index 00000000..40167588 --- /dev/null +++ b/opt/plugin/doctor/collector.yaml @@ -0,0 +1,24 @@ +# collector configuration for doctor verification job +# collect performance data from log files +collector: logfile +# transform collected data into timeline +transformer: timeline +logfiles: + - filename: doctor_consumer.log + # 2016-12-28 03:16:05,630 consumer.py 26 INFO doctor consumer notified at 1482894965.63 + - match: doctor consumer notified at \d+(\.\d+)?$ + checkpoint: notified consumer + - filename: doctor_inspector.log + # 2016-12-28 03:16:05,299 inspector.py 76 INFO event posted at 1482894965.3 + # 2016-12-28 03:16:05,299 inspector.py 56 INFO doctor mark vm(<Server: doctor_vm1>) error at 1482894965.3 + # 2016-12-28 03:16:05,506 inspector.py 66 INFO doctor mark host(overcloud-novacompute-1.ool-virtual1) down at 1482894965.51 + - match: event posted at \d+(\.\d+)?$ + checkpoint: posted event + - match: doctor mark vm\(.*\) error at \d+(\.\d+)?$ + checkpoint: marked VM error + - match: doctor mark host\(.*\) down at \d+(\.\d+)?$ + checkpoint: marked host down + - filename: disable_network.log + # doctor set host down at 1482894965.164096803 + - match: doctor set host down at \d+(\.\d+)?$ + - checkpoint: set host down |