Hello everyone,

We are facing some difficulties on our central server.
The 2007 infrastructure is managing approximatively 100k clients.

We had a few months ago an issue concerning disk space on the server which 
prevented Primary Child sites to send their instructions and status message to 
the Central.
We had a backlog of over 13M files. This backlog has been processed using a 
method where we sent "small" status messages file (<100Ko) and went up in size 
each time it processed a category.

We now "only" have 60k status messages remaining (but files are mainly "big" 
i.e. > 600Ko).
Once we try to make the server process these files, it starts to create a 
backlog (last night generated 600k files of backlog for example).

We investigated and didn't find a proper solution for our issue.

Last option we didn't try was to change settings in the sitectrl.ct0 file to 
raise the number of threads for the SMS_State_System component and change the 
Missing Message Age to go from 2 days to over a week.
But before doing so I'd love to hear some feedback about backlogs and possibly 
someone that changed the control file.

Last but not least, the backup cannot be performed because there are too many 
files to be saved and the backup is timing out every time we try to do so.

Regards,

Jérémy SI HASSEN | InepSol | Architecte Configuration Manager, Poste de Travail 
et Data Protection Manager | +33 6 63 23 22 11 | 
[email protected]<mailto:[email protected]>
[btn_in_20x15]Jsihassen on Linkedin<http://fr.linkedin.com/in/jsihassen>




Reply via email to