New issue
Advanced search Search tips
Note: Color blocks (like or ) mean that a user may not be available. Tooltip shows the reason.

Issue 728803 link

Starred by 1 user

Issue metadata

Status: Fixed
Owner:
Closed: Dec 2017
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 3
Type: Bug



Sign in to add a comment

afe_jobs has at least 4 concepts of timeout

Project Member Reported by akes...@chromium.org, Jun 1 2017

Issue description

Chrome Version: (copy from chrome://version)
OS: (e.g. Win7, OSX 10.9.5, etc...)

What steps will reproduce the problem?
(1)
(2)
(3)

What is the expected result?

What happens instead?

Please use labels and text to provide additional information.


For graphics-related bugs, please copy/paste the contents of the about:gpu
page at the end of this report.
mysql> describe afe_jobs;
+----------------------+--------------+------+-----+---------+----------------+
| Field                | Type         | Null | Key | Default | Extra          |
+----------------------+--------------+------+-----+---------+----------------+
| id                   | int(11)      | NO   | PRI | NULL    | auto_increment |
| owner                | varchar(255) | YES  | MUL | NULL    |                |
| name                 | varchar(255) | YES  | MUL | NULL    |                |
| priority             | int(11)      | YES  |     | NULL    |                |
| control_file         | text         | YES  |     | NULL    |                |
| control_type         | int(11)      | YES  |     | NULL    |                |
| created_on           | datetime     | YES  | MUL | NULL    |                |
| synch_count          | int(11)      | NO   |     | NULL    |                |
| timeout              | int(11)      | NO   |     | NULL    |                |
| run_verify           | tinyint(1)   | YES  |     | 1       |                |
| email_list           | varchar(250) | NO   |     | NULL    |                |
| reboot_before        | smallint(6)  | NO   |     | NULL    |                |
| reboot_after         | smallint(6)  | NO   |     | NULL    |                |
| parse_failed_repair  | tinyint(1)   | NO   |     | 1       |                |
| max_runtime_hrs      | int(11)      | NO   |     | NULL    |                |
| drone_set_id         | int(11)      | YES  | MUL | NULL    |                |
| parameterized_job_id | int(11)      | YES  | MUL | NULL    |                |
| max_runtime_mins     | int(11)      | NO   |     | NULL    |                |
| parent_job_id        | int(11)      | YES  | MUL | NULL    |                |
| test_retry           | int(11)      | NO   |     | 0       |                |
| run_reset            | smallint(6)  | NO   |     | 1       |                |
| timeout_mins         | int(11)      | NO   |     | NULL    |                |
| shard_id             | int(11)      | YES  | MUL | NULL    |                |
| require_ssp          | tinyint(1)   | YES  |     | NULL    |                |
+----------------------+--------------+------+-----+---------+----------------+
24 rows in set (0.00 sec)


timeout, max_runtime_hrs, max_runtime_mins, timeout_mins

I'm pretty sure that at least max_runtime_mins and timeout are used (I see them both referenced in monitor_db_cleanup.py) but not sure of the others.
 
From reading monitor_db_cleanup, my belief is that |timeout| applies to the time we allow an afe_job to live after it's |created_on| (which, I hope, is the insertion time?).

|max_runtime_mins| applies to afe_host_queue_entries and is enforced relative to it's |started_on|. What I don't know is whether |started_on| is the time at which the test started, or at which the pre-test tasks started, and whether the timeout applies to pre-test things.
c#1: That was my reading along with the same uncertainties regarding started_on.

Comment 3 by aut...@google.com, Jun 12 2017

Owner: akes...@chromium.org
1. document this 
2. investigate re-naming obsolete columns (Don thinks it can be done)  as long as the callers don't use the names
Status: Fixed (was: Untriaged)
This bug is not really actionable. Moved its discoveries to doc at https://sites.google.com/a/google.com/chromeos/for-team-members/infrastructure/backend/timeouts-in-autotest

Sign in to add a comment