jc211
/
tractor
forked from goodboy/tractor
1
0
Fork 0
Code Pull Requests Activity
1,147 Commits
211 Branches
0 Tags
9.3 MiB
Commit Graph

1147 Commits (c30cece37a818b07f3c565a057e7f6179eb6f729)

Author SHA1 Message Date
Tyler Goodlet a23afb0bb8 Set channel cancel called flag on cancel requests 2021-12-02 08:18:04 -05:00
Tyler Goodlet 1976e61d1a Add `.drain()` support to msg streams 
Enables "draining" the last set of messages after a channel/stream has
been terminated mostly for the purposes of receiving a final ACK to
a remote cancel command. Also, add an internal `Channel._cancel_called`
flag which can be set by `Portal.cancel_actor()`.
 2021-12-02 08:18:04 -05:00
Tyler Goodlet 0ac3397dbb Only soft-acquire debug lock if a proc was spawned 2021-12-02 08:17:03 -05:00
Tyler Goodlet 62b2867e07 Tweak doc strings 2021-12-02 08:16:49 -05:00
Tyler Goodlet bf6958cdbe Handle cancelled-before-proc-created spawn case 
It's definitely possible to have a nursery spawn task be cancelled
before a `trio.Process` handle is ever returned; we now handle this
case as a cancelled-during-spawn scenario. Zombie collection logic
also is bypassed in this case.
 2021-12-02 08:16:05 -05:00
goodboy d05885d650
Merge pull request #266 from goodboy/faster_daemon_cancels 
Faster graceful daemon cancels
 2021-11-30 09:29:13 -05:00
Tyler Goodlet 77fc705b1f Add nooz 2021-11-29 22:52:19 -05:00
Tyler Goodlet 16a3321a38 Increase timeout for windows.. 2021-11-29 21:52:30 -05:00
Tyler Goodlet 7eb465a699 Graceful cancel actors before hard reaping 2021-11-29 16:03:23 -05:00
Tyler Goodlet 121f7fd844 Draft test that shows a slow daemon cancellation 
Currently if the spawn task is waiting on a daemon actor it is likely in
`await proc.wait()`, however, if the actor nursery is subsequently
cancelled this checkpoint will be abandoned and the hard proc reaping
sequence will execute which results in a up to 3 second wait before
a "hard" system signal is sent to the child.  Ideally such
a cancelled-during-daemon-actor-wait condition is instead handled by
first trying to cancel the remote actor using `Portal.cancel_actor()` (a
"graceful" remote cancel request) which should (presuming normal runtime
operation) result in an immediate collection of the process after normal
actor (remotely triggered) runtime cancellation.
 2021-11-29 16:03:14 -05:00
goodboy ac821bdd94
Merge pull request #264 from goodboy/runinactor_none_result 
Fix `Portal.run_in_actor()` returns `None` result
 2021-11-29 09:21:24 -05:00
Tyler Goodlet f6de7e0afd Factor out msg unwrapping into a func 2021-11-29 08:46:35 -05:00
Tyler Goodlet 0e7234aa68 Cache the return message instead of the value 
Thanks to @richardsheridan for pointing out the limitations of using
*any* kind of value as the result-cached-flag and how it might cause
problems for anyone returning pickled blob-data. This changes the
`Portal` internal result value tracking to stash the full message from
which the value can be retrieved by any `Portal.result()` caller.
The internal change is that `Portal._return_once()` now returns a tuple
of the message *and* its value.
 2021-11-29 07:44:44 -05:00
Tyler Goodlet 83da92d4cb Add nooz 2021-11-28 19:16:47 -05:00
Tyler Goodlet 57e98b25e7 Increase timeout, windows... 2021-11-20 13:08:19 -05:00
Tyler Goodlet 095c94b1d2 Fix `Portal.run_in_actor()` returns `None` bug 
Fixes the issue where if the main remote task returns `None`,
`Portal.result()` would erroneously wait again on the underlying feeder
mem chan since `None` was being used as the cache flag. Instead set the
flag as the channel uid and consider the result collected when set to
anything else (since it would be odd to return that value from a remote
task when you already can read it as part of portal/channel apis).
 2021-11-20 13:02:08 -05:00
Tyler Goodlet f32ccd76aa Add `Portal.result()` is None test case 
This demonstrates a bug where if the remote `.run_in_actor()` task
returns `None` then multiple calls to `Portal.result()` will hang
forever...
 2021-11-20 13:02:08 -05:00
goodboy b527fdbe1a
Merge pull request #263 from goodboy/early_deth_fixes 
Early deth fixes
 2021-11-08 21:23:09 -05:00
Tyler Goodlet 6b0366fe04 Guard against TCP server never started on cancel 2021-11-07 23:49:32 -05:00
Tyler Goodlet dbe5d96d66 Fix missing yield in lock acquirer 2021-11-07 23:48:05 -05:00
goodboy 08fa55a8c3
Merge pull request #260 from goodboy/clusters_and_hot_tips 
Clusters and hot tips
 2021-11-04 12:02:14 -04:00
Tyler Goodlet 546e1b2fa3 Drop unecessary partial 2021-11-04 10:41:25 -04:00
Tyler Goodlet 94a6fefede Add `open_actor_cluster()` eg. to readme 2021-11-02 15:42:19 -04:00
Tyler Goodlet 74f460eba7 Make auto generated child names <parent_name>.<name> 2021-11-02 15:40:15 -04:00
Tyler Goodlet 4cbb8641de Add an `open_actor_cluster()` usage example 2021-11-02 15:37:36 -04:00
Tyler Goodlet 7efb7da300 Start a hot tips for devs doc 2021-11-02 15:08:20 -04:00
goodboy 2c12d39617
Merge pull request #259 from goodboy/alpha3 
Alpha3
 2021-11-02 14:47:28 -04:00
Tyler Goodlet 6a063b3814 Bump release date by a day 2021-11-02 12:59:26 -04:00
Tyler Goodlet 9da1abeecd Super naive attempt to skip 3.10 on windows 2021-11-02 12:19:58 -04:00
Tyler Goodlet 3452e18e6d Toss 3.10 into CI 2021-11-01 14:12:42 -04:00
Tyler Goodlet 8fdc548676 Alpha3 version bump and release notes 2021-11-01 14:02:45 -04:00
goodboy 5dbe8e4b14
Merge pull request #241 from goodboy/trionics 
Trionics
 2021-10-27 13:09:11 -04:00
goodboy 9c13827a14
Merge pull request #256 from overclockworked64/241-news-fragment 
Add a news fragment
 2021-10-27 12:38:15 -04:00
overclockworked64 6da76949fd
Fix the syntax and point to the new package 2021-10-27 17:03:25 +02:00
overclockworked64 49dd230b4f
Add a newline 2021-10-25 20:01:21 +02:00
overclockworked64 c7f59bd483
Add a news fragment 2021-10-25 19:17:42 +02:00
Tyler Goodlet 083b73ad4a Test: don't grab debug lock if not in mode 2021-10-25 10:22:41 -04:00
goodboy 925af28092
Merge pull request #254 from goodboy/graceful_gather 
Change to `gather_contexts()`, use event for graceful exit
 2021-10-25 10:14:01 -04:00
Tyler Goodlet d0f5c7a5e2 Change to `gather_contexts()`, use event for graceful exit 
The api we've made here is actually closer to `asyncio.gather()` but
with opening async context managers instead of funcs. Use another event
to allow for graceful teardown of children on non-cancellation exits
and add a doc string.
 2021-10-24 14:00:01 -04:00
goodboy ebf080b8a2
Merge pull request #253 from overclockworked64/fix-type-annotation 
Fix type annotations
 2021-10-23 19:09:11 -04:00
overclockworked64 50400359b8
Fix type annotations 2021-10-24 00:47:26 +02:00
goodboy 71b8f9f1ea
Merge pull request #252 from goodboy/246_facepalm_backup 
Trionics improvements from @overclockworked64
 2021-10-23 18:10:17 -04:00
overclockworked64 b91adcf38d Get rid of external teardown trigger 2021-10-23 16:17:30 -04:00
overclockworked64 87e3d32992 Get rid of external teardown trigger because #245 resolves the problem 2021-10-23 16:17:30 -04:00
overclockworked64 04895b9d5e Get rid of dumb random uid and use current actor's uid 2021-10-23 16:17:30 -04:00
overclockworked64 b7a4641674 Allow specifying start_method and hard_kill 2021-10-23 16:17:30 -04:00
overclockworked64 c1089dbd95 Add a clustering test 2021-10-23 16:17:30 -04:00
overclockworked64 3130a04c61 Rename a variable and fix type annotations 2021-10-23 16:17:29 -04:00
overclockworked64 6f9229cd09 Cancel nursery 2021-10-23 16:17:29 -04:00
overclockworked64 6e6baf250b Make sure the ID is a str 2021-10-23 16:17:29 -04:00