The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/443524315e1394ce360e6ef7b6564f5b8809677d

commit 443524315e1394ce360e6ef7b6564f5b8809677d
Author: mattcary <mattcary@chromium.org>
Date: Tue May 30 09:11:03 2017

Disable SQLitePersistentCookieStoreTest::UpdateToEncryption.

Breaks cronet K build.

BUG= 727566 
TBR=nharper@chromium.org,mattm@chromium.org

Review-Url: https://codereview.chromium.org/2909923004
Cr-Commit-Position: refs/heads/master@{#475475}

[modify] https://crrev.com/443524315e1394ce360e6ef7b6564f5b8809677d/net/extras/sqlite/sqlite_persistent_cookie_store_unittest.cc

Also affects SQLitePersistentCookieStoreTest.UpdateFromEncryption and SQLitePersistentCookieStoreTest.TestLoadCookiesForKey.

https://luci-milo.appspot.com/buildbot/chromium.linux/Android%20Tests/42303
https://luci-milo.appspot.com/buildbot/chromium.linux/Android%20Tests/42307

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/d9f3bb3bfc9de8fec2defd1df46afc48b23ee673

commit d9f3bb3bfc9de8fec2defd1df46afc48b23ee673
Author: Peter Kasting <pkasting@chromium.org>
Date: Thu Jun 01 04:57:12 2017

Disable more flakily-timing-out tests.

BUG= 727566 
TEST=none
TBR=nharper,mattm

Change-Id: I4755bcab5dda2c908b420b9db933351b2c1e9a56
Reviewed-on: https://chromium-review.googlesource.com/520732
Reviewed-by: Peter Kasting <pkasting@chromium.org>
Commit-Queue: Peter Kasting <pkasting@chromium.org>
Cr-Commit-Position: refs/heads/master@{#476195}
[modify] https://crrev.com/d9f3bb3bfc9de8fec2defd1df46afc48b23ee673/net/extras/sqlite/sqlite_persistent_cookie_store_unittest.cc

The iOS simulator bot has been exhibiting a lot of "shard #0 took too long to complete" running net_unittests.  In checking, the shard is always stopped on one of these SQLitePersistentCookieStore tests (see e.g. https://luci-logdog.appspot.com/v/?s=chromium%2Fbb%2Fchromium.mac%2Fios-simulator%2F18025%2F%2B%2Frecipes%2Fsteps%2Fnet_unittests__iPad_Air_2_iOS_9.0_%2F0%2Fstdout or https://luci-logdog.appspot.com/v/?s=chromium%2Fbb%2Fchromium.mac%2Fios-simulator%2F18011%2F%2B%2Frecipes%2Fsteps%2Fnet_unittests__iPhone_5_iOS_9.0_%2F0%2Fstdout or https://luci-logdog.appspot.com/v/?s=chromium%2Fbb%2Fchromium.mac%2Fios-simulator%2F18010%2F%2B%2Frecipes%2Fsteps%2Fnet_unittests__iPad_Air_2_iOS_9.0_%2F0%2Fstdout ).

I'm disabling the rest of these tests.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/2150a498ab8ba74261d142941533992fb21bd2d8

commit 2150a498ab8ba74261d142941533992fb21bd2d8
Author: Peter Kasting <pkasting@chromium.org>
Date: Thu Jun 01 06:21:48 2017

Disable rest of SQLPersistentCookieStore tests in hopes of greening iOS.

BUG= 727566 
TEST=none
TBR=nharper,mattm

Change-Id: I2cc60aa1b3a15d4eb6a3c027a49e5b487dadb4c2
Reviewed-on: https://chromium-review.googlesource.com/520668
Reviewed-by: Peter Kasting <pkasting@chromium.org>
Commit-Queue: Peter Kasting <pkasting@chromium.org>
Cr-Commit-Position: refs/heads/master@{#476210}
[modify] https://crrev.com/2150a498ab8ba74261d142941533992fb21bd2d8/net/extras/sqlite/sqlite_persistent_cookie_store_unittest.cc

I've prepared https://chromium-review.googlesource.com/c/524238/ to enable the SQLitePersistentCookieStore tests and try to investigate the failures reported on this bug.

The trybot logs for that CL show that the SQLitePersistentCookieStore tests are each running in at most tens of milliseconds on ios-simulator, so I don't see any indicators that these tests take a long time to run and are causing timeouts on their own. The linked failures for ios-simulator don't convincingly show that there is an issue with those tests - 2 of the 3 links have the test run end with no SQLitePersistentCookieStore test pending. I'd guess that these tests are just how far the bot got before something else caused it to timeout.

For the kitkat build, I could use some help reading that flakiness dashboard. What I see is that the test was passing until 5/30, at which point there was no data (I'm assuming the no data is when the test was disabled). I don't see anything there that indicates a flake.

I'm currently inclined to submit the CL to enable these tests.

Re: ios: looks like these may be a more general problem, see  issue 728216 

re: kitkat build: The flakiness dashboard only shows "KitKat Tablet Tester", whereas the linked failure was from "Android Cronet KitKat Builder", so the flakiness dashboard isn't useful there.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/2b38cc5b8a6a09df64273ba7eaff3688aa98d0e9

commit 2b38cc5b8a6a09df64273ba7eaff3688aa98d0e9
Author: Nick Harper <nharper@chromium.org>
Date: Tue Jun 06 19:32:47 2017

Enable SQLitePersistentCookieStore unittests

Bug:  727566 
Change-Id: Ie8f02673fdcedee426c35d54df872101d227f5cc
Reviewed-on: https://chromium-review.googlesource.com/524238
Reviewed-by: Matt Mueller <mattm@chromium.org>
Commit-Queue: Nick Harper <nharper@chromium.org>
Cr-Commit-Position: refs/heads/master@{#477373}
[modify] https://crrev.com/2b38cc5b8a6a09df64273ba7eaff3688aa98d0e9/net/extras/sqlite/sqlite_persistent_cookie_store_unittest.cc

I can reproduce this hang on iOS @HEAD (after the tests were re-enabled), so I'd expect net_unittests to start hanging again at some point.

The root cause on iOS is the same as in  Issue 728216 .  If this is now an iOS-only failure, we should dupe the bug.  If there is still an issue on Android, I'd guess that it has a different root cause.

Revert of https://chromium-review.googlesource.com/c/526633 is in flight because net_unittests just fail again on Android Cronet Data Reduction Proxy Builder when SQLitePersistentCookieStoreTest.TestLoadCookiesForKey timed out at

https://build.chromium.org/p/chromium.android/builders/Android%20Cronet%20Data%20Reduction%20Proxy%20Builder/builds/5109 :

I  549.076s Main  1 failed tests remain.
I  549.077s Main  STARTING TRY #3/3
I  549.077s Main  Will run 1 tests on 4 devices: 05c34aa83444c162, 06596466005cab1a, 0aa6b19043e4b976, 0accc2af43e4affc
I  549.079s TimeoutThread-1-for-run_tests_on_device(05c34aa83444c162)  [host]> /b/build/slave/Android_Cronet_Data_Reduction_Proxy_Builder/build/src/third_party/android_tools/sdk/platform-tools/adb -s 05c34aa83444c162 shell '( pm dump org.chromium.native_test | grep dataDir=; echo "PIPESTATUS: ${PIPESTATUS[@]}" );echo %$?'
I  549.846s run_tests_on_device(05c34aa83444c162)  flags:
I  549.849s TimeoutThread-1-for-run_tests_on_device(05c34aa83444c162)  [host]> /b/build/slave/Android_Cronet_Data_Reduction_Proxy_Builder/build/src/third_party/android_tools/sdk/platform-tools/adb -s 05c34aa83444c162 push /tmp/tmpvrbInG /data/local/tmp/temp_file-d4b84b7f89f8c.sh
I  549.868s TimeoutThread-1-for-run_tests_on_device(05c34aa83444c162)  Large shell command will be run from file: ( p=org.chromium.native_test;am instrument -w -e "$p".NativeTestInstrumentationTestRunner.ShardNanoTimeout 180000000000 -e "$p".NativeTestInstrumentationTestRunner.NativeTestActivity "$p".NativeUnitTestActivity -e "$p".NativeTest.RunInSubThread 1 -e "$p".NativeTestInstrumentationTestRunner.Test SQLitePersistentCookieStoreTest.TestLoadCookiesForKey -e "$p".NativeTestInstrumentationTestRunner.StdoutFile /storage/emulated/legacy/temp_file-bcbbfdaf19dc0.gtest_out "$p"/"$p".NativeTestInstrumentationTestRunner )> ...
I  549.869s TimeoutThread-1-for-run_tests_on_device(05c34aa83444c162)  [host]> /b/build/slave/Android_Cronet_Data_Reduction_Proxy_Builder/build/src/third_party/android_tools/sdk/platform-tools/adb -s 05c34aa83444c162 shell '( sh /data/local/tmp/temp_file-d4b84b7f89f8c.sh );echo %$?'
I  609.902s run_tests_on_device(05c34aa83444c162)  Still working on StartInstrumentation(05c34aa83444c162, org.chromium.native_test/org.chromium.native_test.NativeTestInstrumentationTestRunner, retries=0, extras={'org.chromium.native_test.NativeTestInstrumentationTestRunner.ShardNanoTimeout': 180000000000, 'org.chromium.native_test.NativeTestInstrumentationTestRunner.NativeTestActivity': 'org.chromium.native_test.NativeUnitTestActivity', 'org.chromium.native_test.NativeTest.RunInSubThread': 1, 'org.chromium.native_test.NativeTestInstrumentationTestRunner.Test': 'SQLitePersistentCookieStoreTest.TestLoadCookiesForKey', 'org.chromium.native_test.NativeTestInstrumentationTestRunner.StdoutFile': '/storage/emulated/legacy/temp_file-bcbbfdaf19dc0.gtest_out'}, timeout=180.0, raw=False)
I  669.964s run_tests_on_device(05c34aa83444c162)  Still working on StartInstrumentation(05c34aa83444c162, org.chromium.native_test/org.chromium.native_test.NativeTestInstrumentationTestRunner, retries=0, extras={'org.chromium.native_test.NativeTestInstrumentationTestRunner.ShardNanoTimeout': 180000000000, 'org.chromium.native_test.NativeTestInstrumentationTestRunner.NativeTestActivity': 'org.chromium.native_test.NativeUnitTestActivity', 'org.chromium.native_test.NativeTest.RunInSubThread': 1, 'org.chromium.native_test.NativeTestInstrumentationTestRunner.Test': 'SQLitePersistentCookieStoreTest.TestLoadCookiesForKey', 'org.chromium.native_test.NativeTestInstrumentationTestRunner.StdoutFile': '/storage/emulated/legacy/temp_file-bcbbfdaf19dc0.gtest_out'}, timeout=180.0, raw=False)
C  729.928s run_tests_on_device(05c34aa83444c162)  Timed out. Dumping threads.

Aside: I treated this as an urgent revert as SoM said it was a tree closer, and an automated email was sent out saying the tree was being closed, but the trees still seemed to be open on the waterfalls.. Expected?

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/982137a5fe34bd200b39f5ea17b33071296a3dcd

commit 982137a5fe34bd200b39f5ea17b33071296a3dcd
Author: Anita Woodruff <awdf@chromium.org>
Date: Wed Jun 07 09:59:18 2017

Revert "Enable SQLitePersistentCookieStore unittests"

This reverts commit 2b38cc5b8a6a09df64273ba7eaff3688aa98d0e9.

Reason for revert: Causing net_unittests to fail on chromium.android/Android Cronet Data Reduction Proxy Builder
(Build #5109)


Original change's description:
> Enable SQLitePersistentCookieStore unittests
> 
> Bug:  727566 
> Change-Id: Ie8f02673fdcedee426c35d54df872101d227f5cc
> Reviewed-on: https://chromium-review.googlesource.com/524238
> Reviewed-by: Matt Mueller <mattm@chromium.org>
> Commit-Queue: Nick Harper <nharper@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#477373}

TBR=mattm@chromium.org,nharper@chromium.org
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug:  727566 

Change-Id: Idefadc008d21c45f44bbf6f415cc6b90c41042d2
Reviewed-on: https://chromium-review.googlesource.com/526633
Reviewed-by: Anita Woodruff <awdf@chromium.org>
Commit-Queue: Anita Woodruff <awdf@chromium.org>
Cr-Commit-Position: refs/heads/master@{#477598}
[modify] https://crrev.com/982137a5fe34bd200b39f5ea17b33071296a3dcd/net/extras/sqlite/sqlite_persistent_cookie_store_unittest.cc

We've had almost the entire SQLitePersistentCookieStoreTest suite disabled for months because of failures on iOS?  This seems pretty disturbing to me, particularly given the importance of the feature.

Passing along the bug to Randy, who intrepidly volunteered to take it from me.  I only got as far as trying to repro on the iOS simulator without success (Ran all disabled tests in the fixture for 1000 iterations, was thinking I'd try on Android, and if that failed to fail, would just try re-enabling on trunk).

Based on rohitrao@'s comment in c#9, on iOS this is the same root cause as  issue 728216 , and hence shouldn't be a problem anymore on that platform.  The general disable was a pretty large sledgehammer to deal with the iOS problems.  So I think we should re-enable and disable just those tests on which Android tests fail, and try to figure out what's going wrong with those.  I'll put together a CL to do that (unless I get Android failures from the try bots, in which case I may just try to debug in place).

So, unsurprisingly, re-enabling all of these tests got no failures on the try bots.  As per c#14 I'm focussing on the Android tests, and to make reverting/relanding either, I'm going to first enable everything that was only iOS related, wait to see if that sticks, then re-enable the Android failures and see whether I get something off the main bots that I don't get off the trybots.  Initial re-enable in https://chromium-review.googlesource.com/c/chromium/src/+/755979.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/a8f3655abe3c5cfec3953cf4ddb6a34e246b1694

commit a8f3655abe3c5cfec3953cf4ddb6a34e246b1694
Author: Randy Smith <rdsmith@chromium.org>
Date: Tue Nov 07 20:08:03 2017

Re-enable SQLitePersistentCookieStore tests.

See https://bugs.chromium.org/p/chromium/issues/detail?id=727566#c14
for details.

Bug:  727566 
Cq-Include-Trybots: master.tryserver.chromium.android:android_cronet_tester;master.tryserver.chromium.mac:ios-simulator-cronet
Change-Id: I94271b5608bbd7c5669a5c4c36b4ec0c48a9b9b2
Reviewed-on: https://chromium-review.googlesource.com/755979
Reviewed-by: Matt Menke <mmenke@chromium.org>
Commit-Queue: Randy Smith <rdsmith@chromium.org>
Cr-Commit-Position: refs/heads/master@{#514563}
[modify] https://crrev.com/a8f3655abe3c5cfec3953cf4ddb6a34e246b1694/net/extras/sqlite/sqlite_persistent_cookie_store_unittest.cc

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/0814d806020791cf246ac8e73dbe926619e074ce

commit 0814d806020791cf246ac8e73dbe926619e074ce
Author: Randy Smith <rdsmith@chromium.org>
Date: Fri Nov 10 19:19:46 2017

Re-enable SQLitePersistentCookieStore tests failing on Android.

*Sheriffs*: These tests have previously passed try job runs and failed main bot runs, and hence
this CL may need to be reverted.

Bug:  727566 
Cq-Include-Trybots: master.tryserver.chromium.android:android_cronet_tester;master.tryserver.chromium.mac:ios-simulator-cronet
Change-Id: I058fd2781f5f895e4b26236f2c1e4153a4fb0660
Reviewed-on: https://chromium-review.googlesource.com/764117
Reviewed-by: Matt Menke <mmenke@chromium.org>
Commit-Queue: Randy Smith <rdsmith@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515628}
[modify] https://crrev.com/0814d806020791cf246ac8e73dbe926619e074ce/net/extras/sqlite/sqlite_persistent_cookie_store_unittest.cc

I don't see any new failures in the past week.  I'm going to consider this bug fixed; please re-open (or file a new bug, either's good) if it recurs.