Bug 1429593
Summary: | Metrics collections consistently fail when where last collection date/time is weeks to months prior | |||
---|---|---|---|---|
Product: | Red Hat CloudForms Management Engine | Reporter: | Thomas Hennessy <thenness> | |
Component: | C&U Capacity and Utilization | Assignee: | Greg Blomquist <gblomqui> | |
Status: | CLOSED CURRENTRELEASE | QA Contact: | Tasos Papaioannou <tpapaioa> | |
Severity: | high | Docs Contact: | ||
Priority: | high | |||
Version: | 5.7.0 | CC: | cpelland, dsafford, fweimer, gblomqui, gekis, jdeubel, jhardy, jocarter, lavenel, myoder, obarenbo, saali, simaishi, thenness, tpapaioa | |
Target Milestone: | GA | Keywords: | TestOnly, ZStream | |
Target Release: | 5.8.0 | |||
Hardware: | x86_64 | |||
OS: | Linux | |||
Whiteboard: | c&u | |||
Fixed In Version: | 5.8.0.8 | Doc Type: | If docs needed, set a value | |
Doc Text: | Story Points: | --- | ||
Clone Of: | ||||
: | 1435278 (view as bug list) | Environment: | ||
Last Closed: | 2017-06-12 16:28:00 UTC | Type: | Bug | |
Regression: | --- | Mount Type: | --- | |
Documentation: | --- | CRM: | ||
Verified Versions: | Category: | --- | ||
oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | ||
Cloudforms Team: | CFME Core | Target Upstream Version: | ||
Embargoed: | ||||
Bug Depends On: | ||||
Bug Blocks: | 1435278, 1459598 |
Created attachment 1260466 [details]
one of 5 metrics collection processes failing on same collection target
Created attachment 1260467 [details]
one of 5 metrics collection processes failing on same collection target
Created attachment 1260468 [details]
one of 5 metrics collection processes failing on same collection target
Created attachment 1260469 [details]
one of 5 metrics collection processes failing on same collection target
Thomas, isn't bug 1429617 a duplicate of this one here? Federica, yes the two BZ'z seem to be reporting the same timeout error for the same class.method. feel free to mark one as a duplicate of the other. *** Bug 1429617 has been marked as a duplicate of this bug. *** there had been a bugzilla very similar to this opened and corrected for PitneyBowes some time ago for the VMware environment. I believe that GregB and possibly Keenan were involved in the fixes. For strictly historical captures we already split the captures into days[1]. However, when it's just picking up `last_perf_captured_on`, the calculations are taking the `start_time` and `end_time` at face value. And, `start_time` in this case is equal to `last_perf_capture_on`[2]. It can be changed to use the same date splitting and stacking as is used by the historical captures. [1] https://github.com/ManageIQ/manageiq/blob/d80866895e8f43783d55481f986f656d9589177e/app/models/metric/ci_mixin/capture.rb#L39-L48 [2] https://github.com/ManageIQ/manageiq/blob/d80866895e8f43783d55481f986f656d9589177e/app/models/metric/ci_mixin/capture.rb#L50-L53 New commit detected on ManageIQ/manageiq/master: https://github.com/ManageIQ/manageiq/commit/557132034d200514fd07803b7bc1a74524a996c3 commit 557132034d200514fd07803b7bc1a74524a996c3 Author: Greg Blomquist <gblomqui> AuthorDate: Tue Mar 14 14:54:06 2017 -0400 Commit: Greg Blomquist <gblomqui> CommitDate: Fri Mar 17 10:53:32 2017 -0400 Split metric collections into smaller intervals When a target's `last_perf_capture_on` is far into the past, the next attempt to capture metrics for that target will result in an attempt to gather a potentially huge amount of metrics at one time. This change will break down the intervals between `last_perf_capture_on` and `today` into individual days. This is the same logic that's used when collecting gap metrics. https://bugzilla.redhat.com/show_bug.cgi?id=1429593 app/models/metric/ci_mixin/capture.rb | 35 ++++++++++++++++++----------- spec/models/metric/ci_mixin/capture_spec.rb | 10 +++++++++ 2 files changed, 32 insertions(+), 13 deletions(-) New commit detected on ManageIQ/manageiq/master: https://github.com/ManageIQ/manageiq/commit/f874607fdb104db3e6730473e5be59a51581229b commit f874607fdb104db3e6730473e5be59a51581229b Author: Greg Blomquist <gblomqui> AuthorDate: Fri Mar 17 11:25:24 2017 -0400 Commit: Greg Blomquist <gblomqui> CommitDate: Fri Mar 17 11:25:24 2017 -0400 Test for reversed metric dates Remove a redundant call to `map`, and add a test to make sure that the dates in metrics collection are in a properly reversed order. https://bugzilla.redhat.com/show_bug.cgi?id=1429593 app/models/metric/ci_mixin/capture.rb | 18 ++++++++++-------- spec/models/metric/ci_mixin/capture_spec.rb | 21 ++++++++++++++++++++- 2 files changed, 30 insertions(+), 9 deletions(-) *** Bug 1428566 has been marked as a duplicate of this bug. *** Verified on 5.8.0.13-rc2. |
Created attachment 1260465 [details] one of 5 metrics collection processes failing on same collection target