Lotus Notes ‘replication only client’ problems
We have a problem with our Lotus Notes clients. We are using Notes 7.0.2 servers and also 7.0.2 client in a 'replication only clíent' concept. That means that all clients have a local replica of their mailfile and nobody is working with the mailfile on the server. Mailfiles are replicated to the server.
We have the problem that sometimes, the clients 'hang' for some seconds or even minutes. This oft seems to be at the moment the client is replicating in background.
We have been thinking about view indexing, full text indexing etc. but could not find answers or a solution yet.
We need a solution FAST since users are complaining about it.
Does anybody has experiance on this topic? Help would be appriciated !
Software/Hardware used:
Software/Hardware used:
ASKED:
October 29, 2008 2:50 PM
UPDATED:
November 6, 2008 3:51 PM
RELATED QUESTIONS
Answer Wiki:
Watch the process of one of the clients to see what is pulling all the resources.
When you replicate, it needs to check for all changes since last replication. On large dbs this can take some time. Then it needs to update view indexes and full text indexes.
Does this occur only the first time in the day or each time it replicates?
Also remember that every email sent is replicated twice, once in the mail file and once as a send mail.
To see all answers submitted to the Answer Wiki: View Answer History.
It’s probably view indexing occuring on the database as new mail is replicated. I would open a PMR with IBM to ask for a best practices review of your infrastructure. For instance, maybe your replication schedule is too long, which means a lot of emails come in with each replication and the view has to index a lot. Maybe if you made the schedule shorter, the few new emails would index faster and provide less of an issue.
Just guessing.
Hello Brucelill,
thank you for your reply.
It occurs not only the first time in the day, but more times a day. We are not even sure it is connected to the replication (in background). The problem is that we can not reproduce it. It is difficult to just ‘sit and wait’ until it happens. Reproduction of the problem would be a big help.
If the problem arises only during replication (and the successive view and full text indexing), I would be able to reproduce it. I am going to try that out. Perhaps this is the way to go.
If you have more ideas, let me know. They are welcome.
If you can put the sysinternals processexporer app on the client and when it hangs run it. It will show what task is consuming the resources.
I have found it’s usually the full text indexer or view indexes. Also see if the have googles desktop running, it tends to cause the same issues.
You can the processexplorer from Microsoft here: