I'm trying to better understand how graph node handles subgraph failures and I learned that graph node automatically unfails subgraphs on initialization and during other aspects of the subgraph lifetime.
Based on this comment and this comment it seems that only fatal errors are unfailed while non-fatal errors persist. Is that correct?
Is there a way to query the database and know if a subgraph will be unfailed at some point in the future or is that managed entirely through in memory?
Ninja edit: it seems that non-fatal errors can get wiped too eg here and here where None is passed to the non_fatal_errors vector - might this be a bug?
Are you aware of any blockers that must be resolved before implementing this feature? If so, which? Link to any relevant GitHub issues.
No response
Some information to help us out
[ ] Tick this box if you plan on implementing this feature yourself.
[X] I have searched the issue tracker to make sure this issue is not a duplicate.
Description
I'm trying to better understand how graph node handles subgraph failures and I learned that graph node automatically unfails subgraphs on initialization and during other aspects of the subgraph lifetime.
Based on this comment and this comment it seems that only fatal errors are unfailed while non-fatal errors persist. Is that correct?
Is there a way to query the database and know if a subgraph will be unfailed at some point in the future or is that managed entirely through in memory?
Ninja edit: it seems that non-fatal errors can get wiped too eg here and here where
None
is passed to thenon_fatal_errors
vector - might this be a bug?Are you aware of any blockers that must be resolved before implementing this feature? If so, which? Link to any relevant GitHub issues.
No response
Some information to help us out