[slurm-users] Extreme long db upgrade 16.05.6 -> 17.11.3

Hendryk Bockelmann bockelmann at dkrz.de
Thu Feb 22 03:12:15 MST 2018

Hi Chris,

we were faced with exactly the same problem - update of 16.05.11 to 
17.11.3 took more than 24 hours without finalizing the conversion of job 
table. Finally, we cancelled the process, went back to "old" version 
16.05.11 and restored the database. At that time we had 10.5 million 
jobs in db with a total db size of 11 GB.

Our "solution":
* Activate purge/archive in slurmdbd.conf keeping just 1months of data. 
This reduced the db to 250000 jobs and 600 MB (needed roughly 2 hours 
for initial purge)
* Update to slurm 17.11.3 with reduced db; it took just 15 minutes!


Dr. Hendryk Bockelmann
Wissenschaftliches Rechnen
Abteilung Anwendungen

Deutsches Klimarechenzentrum GmbH (DKRZ)
Bundesstra├če 45 a, D-20146 Hamburg, Germany

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 4973 bytes
Desc: S/MIME Cryptographic Signature
URL: <http://lists.schedmd.com/pipermail/slurm-users/attachments/20180222/18adab68/attachment.bin>

More information about the slurm-users mailing list