Bug 44647 - Broken threading in adtakeover UMC module causes error
Broken threading in adtakeover UMC module causes error
Status: NEW
Product: UCS
Classification: Unclassified
Component: AD Takeover
UCS 5.0
Other Linux
: P5 major (vote)
: ---
Assigned To: Samba maintainers
Samba maintainers
:
Depends on: 44365
Blocks:
  Show dependency treegraph
 
Reported: 2017-05-22 13:13 CEST by Florian Best
Modified: 2020-06-22 16:09 CEST (History)
2 users (show)

See Also:
What kind of report is it?: Bug Report
What type of bug is this?: 6: Setup Problem: Issue for the setup process
Who will be affected by this bug?: 2: Will only affect a few installed domains
How will those affected feel about the bug?: 3: A User would likely not purchase the product
User Pain: 0.206
Enterprise Customer affected?:
School Customer affected?:
ISV affected?:
Waiting Support:
Flags outvoted (downgraded) after PO Review:
Ticket number:
Bug group (optional):
Max CVSS v3 score:


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Florian Best univentionstaff 2017-05-22 13:13:06 CEST
(In reply to Felix Botner from comment #5)
> OK - i still get an "error" but at least the umc server handles this now by 
>      itself (restarts the module)

(In reply to Florian Best from comment #4)
> I have a reliable way to reproduce this error:
> 
> Execute the following command in a shell:
> # while true; do curl
> http://Administrator:Univention.1@localhost/univention/command/adtakeover/
> progress & sleep 0.1; done
> 
> During this command execute two or three times immediately after another the
> following command:
> 
> # curl
> http://Administrator:Univention.1@localhost/univention/command/adtakeover/
> connect -H 'Content-Type: application/json' -d '{"options": {"ad_username":
> "Administrator", "ad_password": "univention", "ip": "10.200.27.112"}}'
> 
> This will cause the UMC-Server crash.
> The bug is triggered by wrong error handling in combinations with threads in
> the adtakeover module. The module process somehow ends itself which causes
> that all request infos are gone but the state machine/socket buffer still
> contains one response which can't be mapped to any request anymore causing
> this exception.

+++ This bug was initially created as a clone of Bug #44365 +++