Need urgent help,
we upgrade our Hub server from 4.6.7a to 6.0.1 CF1 (OS: Windows NT 4 SP6). After a day of normal work suddenly the nserver.exe application begins to use 99% of the CPU time but all server tasks like router or replicator are idle. The server console can be used as if nothing is wrong. When terminating the
server he will shut down all tasks but hangs with 99% CPU time. After a restart of the server he will run for 1 to 2 hours without any problem but then he will again collect all the CPU time.
Anyone having the same problem or any hint what we can do?
Subject: Domino processes using 99% of CPU
I had a similar problem with different processes. Fixed by running fixup and compact on names.nsf and log.nsf.
Subject: Help: nserver.exe using 99% of CPU
When the server gets in this state, run NSD. (go to the data directory and run [ProgramDir]\nsd.exe). Then you can terminate the server normally and type quit at the NSD prompt.
The NSD log file will contain a snapshot of the server in this state. In particular, it will have stack traces of all the nserver.exe threads. You might find that many of the threads were doing something similar. There might be other clues. If you find something of interest, post it here . With any luck …
Subject: Problem solved
Hello,
we have identified the problem which is causing
the nserver.exe to take away our CPU time. A second server was trying to replicate a database with the Hub server. The replica of this database on the Hub server was corrupt. A fixup on that database solved the problem. This massage appeared during the fixup:
“Unread list for user Server2/Domain01 in database e:\notesdat\lotus\domino\data\appl\de\appl84.nsf is damaged and has been deleted: Invalid ID table”
One symptom of the described problem is that all the
Database Server Tasks showed a status “Buffer Manager I/O Thread”.
Thanks for the help