PR - Ticket 50877 - task to run tests of csn generator

[389-ds-bot]

Cloned from Pagure Pull-Request: https://pagure.io/389-ds-base/pull-request/50878

• Created at 2020-02-03 17:13:11 by tbordaz (@tbordaz)
• Merged at 2020-04-22 15:23:57

---

Bug Description:
It exists a test of csn generator (csngen_test()).
It is not called from any function.

Fix Description:
register a task container 'cn=csngen_test,cn=tasks,cn=config"
that calls the test function

Resolves: #3930

Reviewed by: ?

Platforms tested: F29

Flag Day: no

Doc impact: no

[389-ds-bot]

Comment from firstyear (@Firstyear) at 2020-02-04 02:10:58

This code all looks fine, is there a reason not to put it behind -DDEBUG though? IE is there a need for end-users to be able to run this?

[389-ds-bot]

Comment from lkrispen (@elkris) at 2020-02-04 09:48:07

This code all looks fine, is there a reason not to put it behind -DDEBUG though? IE is there a need for end-users to be able to run this?

I think we frequently have customer deployments showing issues we cannot reproduce. Being able to run this test could be useful. I would not document it though.

[389-ds-bot]

Comment from tbordaz (@tbordaz) at 2020-02-04 10:36:51

@Firstyear , @elkris thanks for the review.
From test pov, calling csngen_test could help to detect a regression if csn generator code changes. For example it is easier to simulate remote/offset with this tests than to play with the system time on a test topology. Regarding tests I think some improvements can be done:

• we should create a tool to parse logs can confirm generated CSN always increase
• check generated CSN conforms the csn generator tuning

For detection in customer deployement we also need more work. at the moment it tests adjustment functions with random csn ahead. It would be interesting to see how csn generator behave with adjustment in the past. An other option would be to replay in real time CSN logged in access log.
Running on customer machine it would also be impacted by system time shift.

[389-ds-bot]

Comment from firstyear (@Firstyear) at 2020-02-05 02:18:28

Okay, I'm happy with this :) It may be good to have lib389 tests and functions related to this too to help us automate this - we can even make tools into the cli that don't list on --help, but could create the test if required.

[389-ds-bot]

Comment from tbordaz (@tbordaz) at 2020-04-21 18:38:37

@Firstyear, was it a ack ?

sorry for the delay, the PR got out of my radar but it could now be useful for additional tests with a library playing with the time.

[389-ds-bot]

Comment from firstyear (@Firstyear) at 2020-04-22 01:32:17

Yes it's an ack, but also would be good to have some lib389 tests for it later too to show it works as you want :)

[389-ds-bot]

Comment from tbordaz (@tbordaz) at 2020-04-22 14:02:44

rebased onto 25f46b6931d328e86eaa37185a454010b551360e

[389-ds-bot]

Comment from tbordaz (@tbordaz) at 2020-04-22 15:20:58

rebased onto 88a33cde321bc178a16949494132d1908864ac74

[389-ds-bot]

Comment from tbordaz (@tbordaz) at 2020-04-22 15:22:51

rebased onto 69a389e

[389-ds-bot]

Comment from tbordaz (@tbordaz) at 2020-04-22 15:23:33

Thanks. Updated patch with lib389 and testcase :)

[389-ds-bot]

Comment from tbordaz (@tbordaz) at 2020-04-22 15:23:57

Pull-Request has been merged by tbordaz

[389-ds-bot]

Patch
50878.patch