CheckoutJobStats
Introduction
New cron job from Charles that runs once per hour and collects job info from the Panda monitor, then stores this in a local SQL database.
- The script is at http://repo.mwt2.org/viewvc/JobDB
- The database is called 'panda' on uct2-grid7. You can access this with the login/password '***/***'. So,
mysql -u * r -h uct2-grid7 -p panda
then the job info is all in the table called 'jobs'.
Jobs database (local)
mysql> describe jobs;
+--------------------+------------------+------+-----+---------+-------+
| Field | Type | Null | Key | Default | Extra |
+--------------------+------------------+------+-----+---------+-------+
| PandaID | int(10) unsigned | | PRI | 0 | |
| prodDBlock | varchar(255) | YES | | NULL | |
| destinationDBlock | varchar(255) | YES | | NULL | |
| startTime | datetime | YES | | NULL | |
| endTime | datetime | YES | | NULL | |
| modificationTime | datetime | YES | | NULL | |
| jobStatus | varchar(16) | YES | | NULL | |
| computingSite | varchar(32) | YES | | NULL | |
| attemptNr | int(11) | YES | | NULL | |
| cpuConsumptionTime | int(11) | YES | | NULL | |
| prodSourceLabel | varchar(32) | YES | | NULL | |
| prodUserId | varchar(64) | YES | | NULL | |
| currentPriority | int(11) | YES | | NULL | |
+--------------------+------------------+------+-----+---------+-------+
13 rows in set (0.00 sec)
--
RobGardner - 14 Apr 2009