Skip to content

connector_flow: Fix issue where it tries to update chunk status in a broken transaction#3

Closed
taezaz wants to merge 1 commit into
initOS:8.0-connector_flowfrom
taezaz:8.0-connector_flow
Closed

connector_flow: Fix issue where it tries to update chunk status in a broken transaction#3
taezaz wants to merge 1 commit into
initOS:8.0-connector_flowfrom
taezaz:8.0-connector_flow

Conversation

@taezaz
Copy link
Copy Markdown

@taezaz taezaz commented Jul 5, 2016

This might happen for example if multiple workers try to execute same job at the same time. It may cause an error "could not serialize access due to concurrent update", which will abort the transaction, yet it will still call chunk.write({'state': new_state}) after that.

…ored until end of transaction block" that occurs when trying to update chunk status after a postgresql error, which leaves transaction invalid
@tremlin
Copy link
Copy Markdown

tremlin commented Aug 25, 2016

Let's continue the discussion in #5 . Basically #3, #4, and #5 are all about the same issue.

@codingforfun
Copy link
Copy Markdown

This is the same issue as #4 but I think the solution isn't correct.
I think the extra rollback isn't necessary because the connector itself will rollback the session in this case.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants