Conversation
This comment has been minimized.
This comment has been minimized.
retries require a clonable request so it is now implemented on subgraph::Request
Geal
changed the title
|
I added some configuration options, marked as experimental, so we can move forward on this and at least have some form of retries, before we settle on the policy configuration format. |
o0Ignition0o
left a comment
o0Ignition0o
left a comment
There was a problem hiding this comment.
Can we open a followup issue on documentation? I think I have an intuition on how to customize it, but it s all still fuzzy to me
BrynCooke
left a comment
BrynCooke
left a comment
There was a problem hiding this comment.
Approved, but I do think documentation should be added as part of this PR.
|
did we decide where documentation should be written for experimental features? |
|
@Geal On my side for logging and tracing I wrote the documentation in the related section. For tracing in tracing.mdx for example. And just mark it as experimental |
|
alright, there's docs now. I feel like the traffic shaping doc could benefit from longer explanations of each feature, and maybe a description of their order and interaction |
|
also, as a follow-up, I think I'll add an option on retries, to deactivate them on mutations if necessary (not all mutations are idempotent) |
implements retries for subgraph requests. This uses Finagle's retry buckets algorithm
5 participants
Fix: #338
Fix: #1956