There are green builds. Starting to look at the new update process.

Testing build R68-10635.0.0 on chromeos2-row4-rack7-host10.

updating before the test has failed twice in a row.

dgarrett$update_servohost -i $BUILD $HOST
Staging beaglebone_servo-release/R68-10635.0.0 on 100.115.219.129
Warning: Permanently added '100.115.219.129' (ECDSA) to the list of known hosts.
chromeos-test@100.115.219.129's password: 
Running this command on chromeos2-row4-rack7-host10-servo.cros:
     update_engine_client --update --omaha_url=http://100.115.219.129:8082/update/beaglebone_servo-release/R68-10635.0.0
[0501/171459:INFO:update_engine_client.cc(471)] Forcing an update by setting app_version to ForcedUpdate.
[0501/171459:INFO:update_engine_client.cc(473)] Initiating update check and install.
[0501/171459:INFO:update_engine_client.cc(502)] Waiting for update to complete.
[0501/171944:ERROR:update_engine_client.cc(217)] Update failed, current operation is UPDATE_STATUS_IDLE, last error code is ErrorCode::kPostinstallRunnerError(5)
Connection to chromeos2-row4-rack7-host10-servo.cros.corp.google.com closed.

The update seems to be failing with:


lsb-release inside the new rootfs:
CHROMEOS_RELEASE_APPID={1BB651DD-C762-3FCF-2A66-CEB4C1096BB1}
CHROMEOS_BOARD_APPID={1BB651DD-C762-3FCF-2A66-CEB4C1096BB1}
CHROMEOS_CANARY_APPID={90F229CE-83E2-4FAF-8479-E368A34938B1}
DEVICETYPE=OTHER
CHROMEOS_RELEASE_BUILDER_PATH=beaglebone_servo-release/R68-10635.0.0
GOOGLE_RELEASE=10635.0.0
CHROMEOS_DEVSERVER=
CHROMEOS_RELEASE_BOARD=beaglebone_servo
CHROMEOS_RELEASE_BUILD_NUMBER=10635
CHROMEOS_RELEASE_BRANCH_NUMBER=0
CHROMEOS_RELEASE_CHROME_MILESTONE=68
CHROMEOS_RELEASE_PATCH_NUMBER=0
CHROMEOS_RELEASE_TRACK=dev-channel
CHROMEOS_RELEASE_DESCRIPTION=10635.0.0 (Official Build) dev-channel beaglebone_servo test
CHROMEOS_RELEASE_BUILD_TYPE=Official Build
CHROMEOS_RELEASE_NAME=Chrome OS
CHROMEOS_RELEASE_VERSION=10635.0.0
CHROMEOS_AUSERVER=https://tools.google.com/service/update2

Set boot target to /dev/mmcblk0p5: Partition 5, Slot B
SetImage
KERNEL_CONFIG: console= loglevel=7 init=/sbin/init cros_secure oops=panic panic=-1 root=PARTUUID=%U/PARTNROFF=1 rootwait ro dm_verity.error_behavior=3 dm_verity.max_bios=-1 dm_verity.dev_wait=0 dm="1 vroot none ro 1,0 716800 verity payload=ROOT_DEV hashtree=HASH_DEV hashstart=71680
0 alg=sha1 root_hexdigest=483f05f0e7f3eef09ad9ce575cd57fade5f1895a salt=bef58d87b134cd9385313419989859021a2d6fa8910720dce2308e43e699c090" noinitrd vt.global_cursor_default=0 kern_guid=%U 
Setting up verity.
Finished after 34 seconds.
Clearing network driver boot cache: /var/lib/preload-network-drivers.
Syncing filesystems before changing boot order...
Finished after 2 seconds.
Updating Partition Table Attributes using CgptManager...
Unable to set successful to 0 for kernel 4
PostInstall Failed

I'm going to pick a different servo to test with. That one has been up and running for 80 days (until I rebooted it). If it was getting regular updates, that shouldn't be possible.

I filed http://b/79093879 for that servo, and started again with chromeos2-row4-rack7-host3.

The update test for that servo worked perfectly, but the downgrade post-test is failing.

dgarrett$update_servohost $HOST
Invalid topic stable_version
Staging beaglebone_servo-release/ on 100.115.219.129
Warning: Permanently added '100.115.219.129' (ECDSA) to the list of known hosts.
chromeos-test@100.115.219.129's password: 
update_servohost: failed to stage beaglebone_servo-release/

That downgrade error was caused by  https://crbug.com/838686 . Working around it now.

Testing finished successfully. Servo versions updated.

dgarrett$stable_version -t cros beaglebone_servo $BUILD
Updating  Chrome OS  beaglebone_servo -> R68-10599.0.0 to R68-10635.0.0

dgarrett$stable_version -t cros guado_labstation $BUILD
Updating  Chrome OS  guado_labstation -> R68-10599.0.0 to R68-10635.0.0
Unchanged Firmware   guado_labstation -> Google_Guado.6301.108.4