mirrors/git-annex
2
0
Fork 0
Code Issues Pull requests Projects Releases Packages 1 Wiki Activity Actions

comment and close

Browse source
This commit is contained in:
Joey Hess 2020-06-22 14:38:09 -04:00
parent b904748ee3
commit e2f739b0c8
No known key found for this signature in database
GPG key ID: DB12DB0FF05F8F38
2 changed files with 23 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

2
doc/bugs/testremote_breeds_way_too_many_instances_of_the_externals_remote.mdwn
View file

@ -217,3 +217,5 @@ I also wonder if there could be some unittest added to test that annex stays "le
[[!meta author=yoh]] [[!meta author=yoh]]
[[!tag projects/datalad]] [[!tag projects/datalad]]
> [[fixed|done]] --[[Joey]]

21
doc/bugs/testremote_breeds_way_too_many_instances_of_the_externals_remote/comment_2_daa09ac406dea6cd3ab74fe19d89d956._comment Normal file
View file

@ -0,0 +1,21 @@
[[!comment format=mdwn
username="joey"
subject="""comment 2"""
date="2020-06-22T18:27:30Z"
content="""
> I also wonder if there could be some unittest added to test that annex stays "lean" (in terms of child processes) while using external remotes.
Normally it will only start one per thread, testremote is unique in
generating variants of a remote.
It would be good to somehow catch situations where a lot of child processes
get spawned in general, like the earlier -J8 starting too many
git helper processes. But I don't know how a test suite could detect or
judge that.
Some of the work done for
[[todo/more_extensive_retries_to_mask_transient_failures]]
avoids potential situations where part of git-annex could leak a process
during exception handling. Continuing work on that will probably improve
things further.
"""]]