1190000 - Autophone - PROCESS-CRASH | autophone-s1s2 | java.lang.NullPointerException at android.net.nsd.NsdManager$ServiceHandler.handleMessage(NsdManager.java:327)

Status: RESOLVED DUPLICATE of bug 1188935

(Firefox for Android Graveyard :: General, defect)

Description

[Bob Clary [:bc] (inactive)]

Beginning with https://treeherder.mozilla.org/#/jobs?repo=mozilla-inbound&revision=be21dfb89dc7 there is a persistent crash on Android 4.2. The other devices are lagging but it is at least intermittent crash on Android 4.3, 4.4

PROCESS-CRASH | autophone-s1s2 | java.lang.NullPointerException at android.net.nsd.NsdManager$ServiceHandler.handleMessage(NsdManager.java:327)

PROCESS-CRASH | autophone-s1s2 | application crashed [@ Java_org_mozilla_gecko_GeckoAppShell_reportJavaCrash]

This is very apparent if you look at only the failures starting with 2015-07-31.

https://treeherder.mozilla.org/#/jobs?repo=mozilla-inbound&filter-searchStr=autophone&exclusion_profile=false&filter-resultStatus=testfailed&filter-resultStatus=busted&filter-resultStatus=exception&startdate=2015-07-31&enddate=2015-08-01

I also saw an illegal argument crash related to nsd but lost it in treeherder. I'll find it again and post here.

Comment 1

[Bob Clary [:bc] (inactive)]

Gary, this is the second nsd related crash this week. It is causing autophone to miss a number of measurements. Can we back this out until you figure out the issue?

Comment 2

[Bob Clary [:bc] (inactive)]

The Illegal Argument issue is unrelated. Fwiw, this also occurs in webappstartup.

Comment 3

[Gary Chen [:xeonchen]]

Hi Bob,

Yes, please back it out.
I'm investigating this issue but no solution currently.

Comment 4

[Bob Clary [:bc] (inactive)]

The behavior appeared to get better over the day with regard to the non-4.2 devices. Then the problem appeared to go away completely with the 4.2 devices as well.

The last build where all of the 4.2 tests failed with PROCESS-CRASH | autophone-s1s2 | java.lang.NullPointerException at android.net.nsd.NsdManager$ServiceHandler.handleMessage(NsdManager.java:328)
was

https://treeherder.mozilla.org/#/jobs?repo=mozilla-inbound&revision=12d229711683&exclusion_profile=false&filter-searchStr=autophone 

The next build showed green

https://treeherder.mozilla.org/#/jobs?repo=mozilla-inbound&revision=ffac74e3366c&exclusion_profile=false&filter-searchStr=autophone

This began to show on b2g-inbound after the merge

https://treeherder.mozilla.org/#/jobs?repo=b2g-inbound&revision=6fe323a36bb2&exclusion_profile=false&filter-searchStr=autophone

but may go away when the monday morning merge happens.

Lets not back out just yet.

Gary, does that check in for Bug 1164725 or Bug 1173930 seems like it could have fixed this problem or covered it up?

Comment 5

[Gary Chen [:xeonchen]]

No, I don't think they are related.

In Bug 1188935, I expected this issue would be disappeared when Bug 1180596 is landed, but sadly seems the patch make it persistent crash as you mentioned in this bug.

Bob, is it possible for me to test this on try server? I'm not sure which option to check?

Comment 6

[Bob Clary [:bc] (inactive)]

Yes, you can use

try: -b o -p android-api-9,android-api-11 -u autophone-s1s2,autophone-webapp -t none

When reviewing your results on try, click the square box between the classified count and the text input at the top right to display hidden jobs.

Please be conservative in the number of try jobs you submit since it is very easy to DOS the limited of devices.