1466569 - The connection was reset for partial content (part 2)

Status: NEW

(Core :: Audio/Video: Playback, defect, P3)

Description

[Jean-Yves Avenard [:jya]]

Bug 1450607 still occurs.

From bug 1450607 comment 34

"Unfortunately, mozregression can't generate me a pushlog ("Unable to find enogh data to bisect" error).
I found that the last good build is 2017-05-05 (buildID: 20170505030252) and first bad build is 2017-05-06 (buildID: 20170506030204). I hope this helps you."

the code is racy, bug 1450607 reduce the likelihood of the race to occur by changing the timing.
But there was never a clear explanation on why bug 1450607 did fix it.

Comment 1

[Jean-Yves Avenard [:jya]]

pushlog:
https://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=0b255199db9d6a6f189b89b7906f99155bde3726&tochange=37a5b7f6f101df2eb292b1b6baaf1540c9920e20

likely bug 1347174 which made change on readahead_limit to limit how much data we download ahead of currentTime...

If you set the pref: media.cache_readahead_limit to 999999 and media.cache_resume_threshold to 999999 with current nightly does the issue still occurs?

(unfortunately, bug 1448222 removed those prefs from being able to be easily modified :( (while not documenting it) )

Comment 2

[Nicholas Nethercote [inactive]]

> If you set the pref: media.cache_readahead_limit to 999999 and
> media.cache_resume_threshold to 999999 with current nightly does the issue
> still occurs?
> 
> (unfortunately, bug 1448222 removed those prefs from being able to be easily
> modified :( (while not documenting it) )

They can be changed in about:config just like any other pref.

Comment 3

[Jean-Yves Avenard [:jya]]

they need to be created....

Comment 4

[Camelia Badau [:cbadau], Desktop Test Engineering]

I've tested on Windows 7 x64 and Windows 10 x64 using latest Nightly 62.0a1 (2018-06-06) with media.cache_readahead_limit and media.cache_resume_threshold prefs changed to 999999 and the issue doesn't occur anymore.

Comment 5

[Liz Henry (:lizzard) (relman/hg->git project)]

jya is this still an issue or can we mark it as fixed/wfm?

Comment 6

[Jean-Yves Avenard [:jya]]

well, the problem still exists... so we should keep this open.

Comment 7

[David Durst [:ddurst]]

If this is part 2 of bug 1450607 which was fixed in 61 and 60.1, then this should also be affecting 60ESR.

Comment 8

[David Durst [:ddurst]]

Is this actively being worked on -- should we expect it for 60.2?

Comment 9

[Nils Ohlmeier [:drno]]

It's my understanding that this is not a recent regression. Given the current resource situation nobody is working on this in the short term.

Comment 10

[Liz Henry (:lizzard) (relman/hg->git project)]

Happy to take a patch in nightly; if it seems low risk enough please feel free to request uplift to 65 beta.