[OpenAFS] Re: UDP timeouts
Thu, 5 May 2011 10:22:41 -0500
On Thu, 05 May 2011 16:40:25 +0200
Jaap Winius <firstname.lastname@example.org> wrote:
> Nevertheless, 10-15 minutes is still 20-30x the default value. As an
> alternative solution, could setting something like...
> fs checkservers -interval 10
> ... on the clients be just as effective? Or, even if it is, would that
> simply be a less elegant solution?
'fs checkservers -interval' affects the probing interval for _down_
fileservers. Probing fileservers that are up is always 10 minutes.
And even if you could change the other interval, having every single
client hit every fileserver every 10 seconds doesn't sound good for
performance. However, 1.6 clients and later will "ping" fileservers
every 20 seconds, but in a manner much more lightweight than an RPC