SYMPTOM:
Users attempting to join two Strategy Intelligence Servers in a cluster may find that one or both the server may hang when attempting to login a user. This behavior may also be noticed on Strategy Intelligence Server startup if the server is set to join the cluster on startup and scheduled user sessions are active.
CAUSE:
This issue is caused by non responsive NFS lock manager services (either statd or lockd) on either the NFS server or the client (Strategy Intelligence Server machine). These services are required for proper management of the file lock requests made by the Strategy Intelligence Server for clustered Inbox and cache files.
When attempting to access clustered file resources on login, the Strategy Intelligence Server will lock the internal session manager component to ensure that no file corruption takes place. However this also means that until the file operation completes, users will be unable to login to the server - because of the blocking file operation other users are effectively locked out of the server giving the impression of a hanging server.
ACTION:
Users should ensure that the lockd and statd services are running normally on both the NFS client node and the NFS server.
The following commands are a guide to obtaining the status of these services on some platforms. Strategy Technical Support recommends that any changes to a system be made in consultation with the System Administrator.
All Unix/Linux platforms - use the command
'rpcinfo -p'
to check the ports that the rpc services are running on.
