Hello,
Recently I have noticed unrealisticly big values of DB12 benchmark. Normally it shold be somewhere between 0 and 45. But I noticed that during this summer we had pilots with jobs that recorded values up to 2250.
So far it is observed only on one computing resource with AlmaLinux and CPU Model Intel Gold 6242R.
Here are zoomed version from our local monitoring. Every dot is a job, color is CPU model, X is DB12 value, Y is job walltime.
I have checked the databases:
mysql> use JobDB;
mysql> select JobID, CONVERT(Value USING utf8) from JobParameters where Name="CPUNormalizationFactor" and CONVERT(Value USING utf8)>45;
+---------+---------------------------+
| JobID | CONVERT(Value USING utf8) |
+---------+---------------------------+
| 4298295 | 60.7 |
| 4298419 | 45.8 |
| 4298478 | 48.8 |
| 4298494 | 48.8 |
+---------+---------------------------+
4 rows in set (0.48 sec)
Unfortunately, jobs with the craziest DB12 benchmark was cleand from DIRAC and there are only resent jobs with wrong DB12 benchmark. But I was able to grep DIRAC Pilot logs on the computing resource to get this:
2024-06-28T08:38:56.134271Z INFO [ConfigureCPURequirements] Command ConfigureCPURequirements instantiated from pilotCommands
2024-06-28T08:38:56.135707Z INFO [ConfigureCPURequirements] Executing command dirac-wms-cpu-normalization -U -R pilot.cfg --cfg pilot.cfg -d
Computing benchmark measurements on 1 processor(s)...
Applying a correction on the CPU power: 1.0
Estimated CPU power is 2272.7 HS06
Configuration file dumped 'pilot.cfg'
2024-06-28T08:39:17.640561Z INFO [ConfigureCPURequirements] Current normalized CPU as determined by 'dirac-wms-cpu-normalization' is 2272.700000
Does anybody saw with this before?
Hello,
Recently I have noticed unrealisticly big values of DB12 benchmark. Normally it shold be somewhere between 0 and 45. But I noticed that during this summer we had pilots with jobs that recorded values up to 2250.
So far it is observed only on one computing resource with AlmaLinux and CPU Model Intel Gold 6242R.
Here are zoomed version from our local monitoring. Every dot is a job, color is CPU model, X is DB12 value, Y is job walltime.
I have checked the databases:
Unfortunately, jobs with the craziest DB12 benchmark was cleand from DIRAC and there are only resent jobs with wrong DB12 benchmark. But I was able to grep DIRAC Pilot logs on the computing resource to get this:
Does anybody saw with this before?