The following revision refers to this bug:
  https://chrome-internal.googlesource.com/chrome-golo/chrome-golo/+/b0a2951da032552149b8f22b7dcb8758af9e6a56

commit b0a2951da032552149b8f22b7dcb8758af9e6a56
Author: John Weathersby <johnw@google.com>
Date: Thu May 10 19:31:08 2018

The following revision refers to this bug:
  https://chrome-internal.googlesource.com/infra/puppet/+/c627ddd7203294876dd6acbca225f0a650b203f6

commit c627ddd7203294876dd6acbca225f0a650b203f6
Author: John Weathersby <johnw@google.com>
Date: Thu May 10 23:29:14 2018

The following revision refers to this bug:
  https://chrome-internal.googlesource.com/chrome-golo/chrome-golo/+/214122ec535abb83fe229fd4c8465cc414644d47

commit 214122ec535abb83fe229fd4c8465cc414644d47
Author: John Weathersby <johnw@google.com>
Date: Fri May 11 00:00:10 2018

All devices migrated from build{112..114}-b1 -> build{202..204}-b7

Docker changes synced, and devices are seen in private swarming now. pool:unassigned

The following revision refers to this bug:
  https://chrome-internal.googlesource.com/infradata/config/+/171dbc755c528108b062f2b3863b61520fc6bcf6

commit 171dbc755c528108b062f2b3863b61520fc6bcf6
Author: Nghia Nguyen <nednguyen@google.com>
Date: Fri May 11 12:39:16 2018

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/tools/build/+/3fbbeee45018a68f2aef9a747df77fcadc9e8a5e

commit 3fbbeee45018a68f2aef9a747df77fcadc9e8a5e
Author: Nghia Nguyen <nednguyen@google.com>
Date: Fri May 11 12:52:55 2018

Convert 'Nexus 6 Webview Perf' to use private swarming server

Bug:841492
Change-Id: If9b934c7547298db8b18bc6b5d2689eebdcfdad8
TBR=eyaich@chromium.org
Reviewed-on: https://chromium-review.googlesource.com/1055468
Reviewed-by: Ned Nguyen <nednguyen@google.com>
Commit-Queue: Ned Nguyen <nednguyen@google.com>

[modify] https://crrev.com/3fbbeee45018a68f2aef9a747df77fcadc9e8a5e/scripts/slave/recipe_modules/chromium_tests/chromium_perf.py

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/34e65e84f06e9e7401447ebb6d1265c5646def1c

commit 34e65e84f06e9e7401447ebb6d1265c5646def1c
Author: Ned Nguyen <nednguyen@google.com>
Date: Fri May 11 12:53:49 2018

Update swarming dimension of 'Android Nexus6 WebView Perf'

NOTRY=true  # Test covered by PRESUBMIT
TBR=eyaich@chromium.org

Bug:  841492 
Change-Id: Id93784c4f9e91e665c8f251349cacd846560ef75
Reviewed-on: https://chromium-review.googlesource.com/1055469
Reviewed-by: Ned Nguyen <nednguyen@google.com>
Commit-Queue: Ned Nguyen <nednguyen@google.com>
Cr-Commit-Position: refs/heads/master@{#557852}
[modify] https://crrev.com/34e65e84f06e9e7401447ebb6d1265c5646def1c/testing/buildbot/chromium.perf.json
[modify] https://crrev.com/34e65e84f06e9e7401447ebb6d1265c5646def1c/tools/perf/core/benchmark_sharding_map.json
[modify] https://crrev.com/34e65e84f06e9e7401447ebb6d1265c5646def1c/tools/perf/core/perf_data_generator.py

John: any idea why a bunch of devices are not shown in chrome-swarming?

https://chrome-swarming.appspot.com/botlist?c=id&c=os&c=task&c=status&f=os%3AAndroid&f=id%3Abuild204-b1--device5&f=pool%3Achrome.tests.perf-webview&l=100&s=id%3Aasc

Looks like I misconfigured the id. The bots ids in #8 was build{202..204}-b1, whereas they're supposed to be build{202..204}-b7

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/99a9a5d575c8385677f15f0a369dcf3f69566de6

commit 99a9a5d575c8385677f15f0a369dcf3f69566de6
Author: Ned Nguyen <nednguyen@google.com>
Date: Fri May 11 23:31:08 2018

Fix device id of 'Android Nexus6 WebView Perf' bots

NOTRY=true  # test by PRESUBMIT
TBR=eyaich@chromium.org

Bug:  841492 
Change-Id: I1b8bf1c05ea18f86ddbf676c24eeac6637a97348
Reviewed-on: https://chromium-review.googlesource.com/1056299
Reviewed-by: Ned Nguyen <nednguyen@google.com>
Commit-Queue: Ned Nguyen <nednguyen@google.com>
Cr-Commit-Position: refs/heads/master@{#558059}
[modify] https://crrev.com/99a9a5d575c8385677f15f0a369dcf3f69566de6/testing/buildbot/chromium.perf.json
[modify] https://crrev.com/99a9a5d575c8385677f15f0a369dcf3f69566de6/tools/perf/core/benchmark_sharding_map.json
[modify] https://crrev.com/99a9a5d575c8385677f15f0a369dcf3f69566de6/tools/perf/core/perf_data_generator.py

John: somehow a few devices are shown up as dead (e.g: build190-b7--device1) & some are shown up as Ubuntu (e.g:build202-b7--device2). Can you check?

https://chrome-swarming.appspot.com/botlist?c=id&c=os&c=task&c=status&f=pool%3Achrome.tests.perf-webview&l=100&q=poo&s=id%3Aasc

Yeah, these seem like localized problems. One host won't even run lsusb, and the other seems to have gone down. We'll investigate these. Thanks.

build190-b7 had a bad usb isolator which locked up the whole chain. This has been replaced.

build202-b7 needed a reboot to clear usb issues as well.

USB device ZX1G422HK6, port num: 1
USB device ZX1G22KC78, port num: 2
USB device ZX1G22LTRP, port num: 3
USB device ZX1G22KGFM, port num: 4
USB device ZX1G22KNX2, port num: 5
USB device ZX1G22L2NR, port num: 6
USB device NP5A2N0156, port num: 7

USB device 01e1781326604096, port num: 1
USB device 008671d42589ad4c, port num: 2
USB device 01e0a954cc6738c6, port num: 3
USB device 0200121bc2432b31, port num: 4
USB device 01dc418a0c9cc4e8, port num: 5
USB device 01e06d0fccaf2eca, port num: 6
USB device 01e131841ca7cc17, port num: 7

Seeing some weirdo usb failures on build190-b7, probably causing the quarantine of devices 2-7. I'll try resetting things. 

build190-b7 looks better now. Though it's taking an awful long time to scan for battors:
"Fetched battor serials in 20.02s."

If this happens again, I may have to dig into that.

Wait, spoke too soon. All but one phone on that host is up:
https://chrome-swarming.appspot.com/bot?id=build190-b7--device4

I think the problem is the battor-scanning issue. We wrap docker's usb interactions around a mutex. Each thread has a timeout of 1min, 1 thread for each device, 7 threads in total, and if scanning battors takes ~20s, a few of those threads will end up timing out. When a timeout happens, a container/swarming-bot ends up not getting a device... I'll fix this when I get a chance.