Hi,
I have a rare issue with imports:
I'm importing a zip file (50MB ~ 50 pdf files), in production environment this operation takes 15 minutes.
It's a cluster with 24GB of RAM each node (16GB of JVM) and a repository with about 1,5 millions of documents.
Monitoring threads and use of servers with JConsole we have about 25% - 35% of resources using.
Same operation in development environment (Same zip file and same operation) takes 45 seconds.
It's a cluster with 8 GB of RAM each node (4GB of JVM) and a repository with 1,3 millions of documents.
Version of Alfresco is 3.4.10
Is there an explanation about the differences of times?
We detected a missmatch in production between indexes of each node of the cluster, in one node, we detected documents and spaces without index. Can this be a explanation of the issue? What other reasons can cause this delay?
Thanks in advance,
Kind Regards