[Bug 260764] Feature request: Control over NFS bg timeout
- Reply: bugzilla-noreply_a_freebsd.org: "[Bug 260764] Feature request: Control over NFS bg timeout"
- Reply: bugzilla-noreply_a_freebsd.org: "[Bug 260764] nfs: Add control over bg timeout"
- Reply: bugzilla-noreply_a_freebsd.org: "[Bug 260764] nfs: Add control over bg timeout"
- Reply: bugzilla-noreply_a_freebsd.org: "[Bug 260764] nfs: Add control over bg timeout"
- Go to: [ bottom of page ] [ top of archives ] [ this month ]
Date: Tue, 28 Dec 2021 17:18:33 UTC
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=260764
Bug ID: 260764
Summary: Feature request: Control over NFS bg timeout
Product: Base System
Version: Unspecified
Hardware: Any
OS: Any
Status: New
Severity: Affects Some People
Priority: ---
Component: kern
Assignee: bugs@FreeBSD.org
Reporter: jwb@freebsd.org
Scenario:
HPC clusters have a head node responsible for managing resources on compute
nodes and often serving as a login node, from which jobs are submitted.
The head node is usually the first node booted since it runs services on which
other nodes depend, but also needs access to file servers, which are booted
later. Mounting file servers with the "bg" option in fstab allows the head
node to complete a boot while other nodes are down, but only after a 60+ second
delay as it waits for the mount to time out.
I have worked around this by using "noauto" instead of "bg", and using a cron
job to mount file servers. This avoids delays in booting, but is messy. It
would be cleaner if there were a way to reduce the timeout for mounting with
"bg" to a few seconds when the server is not responding.
This would also be helpful in other situations where NFS servers need to be
cross mounted.
--
You are receiving this mail because:
You are the assignee for the bug.