This is an old thread (but still the best search match for the error message).
Several other people pointed at NFS, so it got me to look.
Indeed what appear to happen (for me at least) was ....
The folders being imapped, some of them were on a NFSv3 mount. The NFS server had been shutdown.
The imap process appeared to hang (presumably trying to read from the defunct NFS mount point) this
held the lock , so other imap processes (possibly not using NFS) hung awaiting the lock.
The "fix" was to bing the NFS server back online ...and as it turned out restart a number of things
like autofs and dovecot (Normally NFS clients just carry on, so the timeout must have affected things)
My long-term fix came from:
https://docs.oracle.com/cd/E19620-01/805-5233/z400163399f/index.html
Namely to add -soft in /etc/auto_master
This should cause a "fail" instead of a hang.
--
Graeme