Bug 158646

Summary:	[Yosemite Release WK2] imported/w3c/IndexedDB-private-browsing/idbfactory_open9.html is a flaky timeout
Product:	WebKit	Reporter:	Ryan Haddad <ryanhaddad>
Component:	New Bugs	Assignee:	Nobody <webkit-unassigned>
Status:	NEW ---
Severity:	Normal	CC:	ap, beidson, jbedard
Priority:	P2
Version:	WebKit Nightly Build
Hardware:	Unspecified
OS:	Unspecified
See Also:	https://bugs.webkit.org/show_bug.cgi?id=159827
Bug Depends on:
Bug Blocks:	154968

Description Ryan Haddad 2016-06-10 17:10:37 PDT

imported/w3c/IndexedDB-private-browsing/idbfactory_open9.html is a flaky timeout

<https://build.webkit.org/builders/Apple%20Yosemite%20Release%20WK2%20(Tests)/builds/15220>

<https://webkit-test-results.webkit.org/dashboards/flakiness_dashboard.html#showAllRuns=true&tests=imported%2Fw3c%2FIndexedDB-private-browsing%2Fidbfactory_open9.html>

Comment 1 Ryan Haddad 2016-06-10 17:10:58 PDT

Seems to time out alongside imported/w3c/IndexedDB-private-browsing/idbfactory_open9.html

Comment 2 Ryan Haddad 2016-06-10 17:11:49 PDT

(In reply to comment #1)
> Seems to time out alongside
> imported/w3c/IndexedDB-private-browsing/idbfactory_open9.html

By that, I meant imported/blink/storage/indexeddb/blob-delete-objectstore-db.html

Comment 3 Alexey Proskuryakov 2016-06-11 16:19:49 PDT

I wonder if these timeouts are misreported crashes, like in bug 158665.

Comment 4 Brady Eidson 2016-06-11 17:57:16 PDT

(In reply to comment #3)
> I wonder if these timeouts are misreported crashes, like in bug 158665.

I would love a conclusion on that, as some of these IDB timeouts are beyond mysterious - How do we find out for sure?

Comment 5 Alexey Proskuryakov 2016-06-11 21:41:55 PDT

I checked the bot where the build quoted by Ryan took place, and there is no crash stored there that would be relevant. So chances are that it's a true timeout.

Perhaps we could temporarily enable sampling on timeout, or even take a spindump. I'd r+ a patch for that.

Comment 6 Brady Eidson 2016-06-11 21:45:13 PDT

(In reply to comment #5)
> I checked the bot where the build quoted by Ryan took place, and there is no
> crash stored there that would be relevant. So chances are that it's a true
> timeout.
> 
> Perhaps we could temporarily enable sampling on timeout, or even take a
> spindump. I'd r+ a patch for that.

I would r+ one, too, but I certainly don't know how to implement one myself.

Comment 7 Alexey Proskuryakov 2016-06-11 22:01:38 PDT

Whoops, this timeout in particular is not a freeze (at least not in WebKitTestRunner or WebContent). It has output: https://build.webkit.org/results/Apple%20Yosemite%20Release%20WK2%20(Tests)/r201935%20(15220)/imported/w3c/IndexedDB-private-browsing/idbfactory_open9-pretty-diff.html

Enabling sample on WKTR is easy (we'd just need to remove --no-sample from arguments), but this is a Databases process freeze at best, and more likely not a freeze at all.

Comment 8 Brady Eidson 2016-07-04 21:44:57 PDT

(In reply to comment #7)
> Whoops, this timeout in particular is not a freeze (at least not in
> WebKitTestRunner or WebContent). It has output:
> https://build.webkit.org/results/Apple%20Yosemite%20Release%20WK2%20(Tests)/
> r201935%20(15220)/imported/w3c/IndexedDB-private-browsing/idbfactory_open9-
> pretty-diff.html
> 
> Enabling sample on WKTR is easy (we'd just need to remove --no-sample from
> arguments), but this is a Databases process freeze at best, and more likely
> not a freeze at all.

Confirming, this cannot be a database process freeze (or crash, or hang), because this is a private browsing test: Database process is not involved at all.