MonitoringInvalidFieldTypeError: Metric "...resolve_devserver" was given field "devserver" with invalid value |
||
Issue description
11/29 10:19:30.972 ERROR| interface:0166| Automatic monitoring flush failed.
Traceback (most recent call last):
File "/usr/local/autotest/site-packages/chromite/third_party/infra_libs/ts_mon/common/interface.py", line 164, in _flush_and_log_exceptions
flush()
File "/usr/local/autotest/site-packages/chromite/third_party/infra_libs/ts_mon/common/interface.py", line 109, in flush
metric.serialize_to(proto, start_time, fields, value, target)
File "/usr/local/autotest/site-packages/chromite/third_party/infra_libs/ts_mon/common/metrics.py", line 115, in serialize_to
self._populate_fields(metric_pb, fields)
File "/usr/local/autotest/site-packages/chromite/third_party/infra_libs/ts_mon/common/metrics.py", line 142, in _populate_fields
raise errors.MonitoringInvalidFieldTypeError(self._name, key, value)
MonitoringInvalidFieldTypeError: Metric "chrome/infra/chromeos/autotest/resolve_devserver" was given field "devserver" with invalid value "<autotest_lib.client.common_lib.cros.dev_server.AndroidBuildServer object at 0x7f28d8243e50>" (<class 'autotest_lib.client.common_lib.cros.dev_server.AndroidBuildServer'>).
Found in suite scheduler log.
,
Nov 29 2016
suite scheduler needs to be restarted to pick up the change. It was restarted before the revert was pushed to the lab. I just restarted it. It should be fixed now. Will check in an hour or so.
,
Nov 29 2016
Can we restart suite scheduler on push? This kind of limbo state is confusing.
,
Nov 29 2016
suite scheduler is not 100% safe for restart, it might drop certain suite run during restart. Although it's rare case (it used to happen more because it ran way slower), we don't want to take the risk. suite scheduler code is rather stable nowadays. And xixuanw is working on a new design to replace it.
,
Nov 29 2016
|
||
►
Sign in to add a comment |
||
Comment 1 by ayatane@chromium.org
, Nov 29 2016