pikers
/
piker
4
1
Fork 0
Code Issues 11 Pull Requests 12 Packages Projects Releases Wiki Activity

stop_is_oec: expect trio.EndOfChannel as graceful stream shutdown #37

Closed
goodboy wants to merge 0 commits from stop_is_oec into main
pull from: stop_is_oec
merge into: pikers:main
Conversation 3 Commits - Files Changed -
goodboy commented 2025-02-20 16:32:25 +00:00
Owner

This should likely be synced with tractor’s new typed-IPC-msging either via dev branch or (ideally) a release!

Change .data._sampling broadcast error handling to include trio.EndOfChannel in the Sampler.bcast_errors: set since a tractor.MsgStream termination msg (Stop) now instead translates to that raised in task.

The main purpose here is to avoid crashing the sampler broadcast task in such graceful peer-disconnect conditions.

This should likely be synced with `tractor`'s new [typed-IPC-msging](https://pikers.dev/goodboy/tractor/pulls/7) either via dev branch or (ideally) a release! --- Change `.data._sampling` broadcast error handling to include `trio.EndOfChannel` in the `Sampler.bcast_errors: set` since a `tractor.MsgStream` termination msg (`Stop`) now instead translates to that raised in task. The main purpose here is to avoid crashing the sampler broadcast task in such graceful peer-disconnect conditions.
goodboy added 3 commits 2025-02-20 16:32:25 +00:00
6e495511ff Suppress `trio.EndOfChannel`s raised by remote peer 
Since now `tractor` will raise this native `trio`-exc translated from
a `Stop` msg when the peer gracefully terminates a `tractor.MsgStream`.
Just `info()` log in such cases versus continuing to warn for the
others.
a795108041 Group bcast errors as `Sampler.bcast_errors` 
A new class var `tuple[Exception]` such that the err set can be reffed
externally as needed for catching other similar pub-sub/IPC failures in
other (related) real-time sub-systems.

Also added some now-masked logging for debugging live-feed stream reading
issues that should ONLY be used for debugging since they'll greatly
degrade HFT perf. Used the new `log.mk_repr()` stuff (that one day we
should prolly pull from `modden` as a dep) for pretty console emissions.
389c746223 Catch using `Sampler.bcast_errors` where possible 
In all other possible IPC disconnect handling blocks. Also more
comprehensive typing throughout `uniform_rate_send()`.
goodboy force-pushed stop_is_oec from 389c746223 to 49841f5b91 2025-02-21 21:25:07 +00:00 Compare
goodboy changed title from stop_is_oec: expect `trio.EndOfChannel` as grace stream shutdown to stop_is_oec: expect `trio.EndOfChannel` as graceful stream shutdown 2025-02-21 22:12:33 +00:00
goodboy commented 2025-04-24 14:48:41 +00:00
Poster
Owner

The commit 3ff0a8674 from the branch decimal_prices_thru_ems should be cherried to this as well

The commit 3ff0a8674 from the branch `decimal_prices_thru_ems` should be cherried to this as well
goodboy changed title from stop_is_oec: expect `trio.EndOfChannel` as graceful stream shutdown to stop_is_oec: expect `trio.EndOfChannel` as graceful stream shutdown 2025-12-31 08:16:37 +00:00
goodboy changed target branch from jsonrpc_err_in_rent to main 2025-12-31 08:16:37 +00:00
goodboy commented 2025-12-31 19:31:57 +00:00
Poster
Owner

Ok i think the list of non-conflicting cherries and/or TO-DROPs is,

hopefully that’s about right as it’s the change-set on .data/_sampling i see on my workin branch.

i think a rebase of gitea/decimal_prices_thru_ems onto this branch should clarify?

Ok i think the list of non-conflicting cherries and/or TO-DROPs is, - 3ff0a86741836b61c32f0fd0be1b2b227040ef00 - 94caa248e7 hopefully that's about right as it's the change-set on `.data/_sampling` i see on my workin branch. i think a rebase of `gitea/decimal_prices_thru_ems` onto this branch should clarify?
goodboy commented 2026-01-01 19:43:16 +00:00
Poster
Owner

Ok i’ve rebased various downstream branches onto this, going to close in favor of #52 which has the typo in the branch name corrected Bp

eoc = end-of-channel, i.e. trio.EndOfChannel, where “stop is eoc” signifies that stream stopping can be signalled by an eoc as well.

Ok i've rebased various downstream branches onto this, going to close in favor of #52 which has the typo in the branch name corrected Bp eoc = end-of-channel, i.e. `trio.EndOfChannel`, where "stop is eoc" signifies that stream stopping can be signalled by an eoc **as well**.
goodboy closed this pull request 2026-01-01 19:43:16 +00:00

Pull request closed

This pull request cannot be reopened because the branch was deleted.
Sign in to join this conversation.
Reviewers
No reviewers
Labels
Clear labels
No items
No Label
Milestone
Clear milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
No Assignees
1 Participants
Notifications
Due Date
The due date is invalid or out of range. Please use the format 'yyyy-mm-dd'.

No due date set.

Dependencies

No dependencies set.

Reference: pikers/piker#37
There is no content yet.
Delete Branch "stop_is_oec"

Deleting a branch is permanent. Although the deleted branch may exist for a short time before cleaning up, in most cases it CANNOT be undone. Continue?