From 9677814e5d09b6a42c9ddf232eff3b0182d9f433 Mon Sep 17 00:00:00 2001 From: yarikoptic Date: Mon, 12 Oct 2020 16:00:32 +0000 Subject: [PATCH] Added a comment --- ...ent_6_58997665e5aa0a237a3ec90efca92891._comment | 14 ++++++++++++++ 1 file changed, 14 insertions(+) create mode 100644 doc/bugs/some_annex_addurl_--fast_--with-files_--json_--json-error-messages_--batch__do_not_quit/comment_6_58997665e5aa0a237a3ec90efca92891._comment diff --git a/doc/bugs/some_annex_addurl_--fast_--with-files_--json_--json-error-messages_--batch__do_not_quit/comment_6_58997665e5aa0a237a3ec90efca92891._comment b/doc/bugs/some_annex_addurl_--fast_--with-files_--json_--json-error-messages_--batch__do_not_quit/comment_6_58997665e5aa0a237a3ec90efca92891._comment new file mode 100644 index 0000000000..70ecee7b9a --- /dev/null +++ b/doc/bugs/some_annex_addurl_--fast_--with-files_--json_--json-error-messages_--batch__do_not_quit/comment_6_58997665e5aa0a237a3ec90efca92891._comment @@ -0,0 +1,14 @@ +[[!comment format=mdwn + username="yarikoptic" + avatar="http://cdn.libravatar.org/avatar/f11e9c84cb18d26a1748c33b48c924b4" + subject="comment 6" + date="2020-10-12T16:00:31Z" + content=""" +Thank you Joey for the analysis! I am not 100% certain why that nose fork process came about. May be nose (testing framework) causes a fork if some processes are still lingering after a test completes... I guess something for us to keep in mind + +Also I +- [identified the change has caused the issue to disappear](https://github.com/datalad/datalad/pull/5013#issuecomment-707189474) while looking through recent datalad changes. Not yet sure if that was anyhow a \"fix\" but at least I do not observe it any longer +- reproduced the issue with elderly 7.20190819+git2-g908476a9b-1~ndall+1, so it confirms that it is not a regression within git-annex (due to some flaky run I thought about possible regression). + +I think you can close it, and we (datalad) should keep an eye on nose and real-life if any stalls... so far I only observed datalad abandoning `git-annex` process without waiting long enough (like in the original case I reported here) when it could be just taking longer to complete indeed. +"""]]