For some reason our CMS has started to publish really slow every now and then without any apparent reason so for example full publishes go from 1 hour and 20 minutes to more than 4 hours.
Any idea about what I can monitor/log to identify the reason? I changed the logs settings to log all the SQL queries but the response times do seem OK so apparently the DB is not the problem.
I did notice though that the sum of assembling all times is more or less the same both for the slow and quick publishes ( calculated through the following query: SELECT
SUM(ELAPSED) FROM PSX_PUBLICATION_DOC WHERE STATUS_ID = 53218), so it seems that there is something I’m missing here.
Any thoughts or ideas on this would be appreciated.
I’m not sure I’m following you, the SQL query bellow is just to know how long did it take to Rhythmyx to assembly all the items involved on a specific publish, you can see same information when you click on the logs of any publish but you can’t see the total its displayed on a per item basis.
Usually when I’ve seen random slowdowns like this it is because there is a backup going on either on the database, the server that you are publishing too, or on the RX server itself. It could also be a large network operation on the same segment that is slowing down traffic in general.
Is it a file system publisher?
-n
[QUOTE=javier;19113]Hi guys,
For some reason our CMS has started to publish really slow every now and then without any apparent reason so for example full publishes go from 1 hour and 20 minutes to more than 4 hours.
Any idea about what I can monitor/log to identify the reason? I changed the logs settings to log all the SQL queries but the response times do seem OK so apparently the DB is not the problem.
I did notice though that the sum of assembling all times is more or less the same both for the slow and quick publishes ( calculated through the following query: SELECT
SUM(ELAPSED) FROM PSX_PUBLICATION_DOC WHERE STATUS_ID = 53218), so it seems that there is something I’m missing here.
Any thoughts or ideas on this would be appreciated.
We just fixed this last Friday, the problem seemed to be caused by an excess of stored publications logs (which it doesn’t explain why just sometimes the system manage to publish normally).
As soon as the problem started we did not want to delete any publishing logs to try to figure out what was going on when in fact we were just making the problem worst.
Thanks for all your responses and thoughts on this.