[OpenAFS] Doubts about OpenAFS implementation in a company
Thu, 19 May 2011 08:05:20 +0200
So now we're getting somewhere :-)
In fact we have 6 (or 7) DB servers. I remember that at one point my
colleagues were surprised by "8 DB servers limit" - because up to this
point it was a common practice to install full set of daemons on every
server. Now we've got multiple file servers in Polish offices, and
usually just a single DB server in each.
The reason is quite simple: although all of our links are redundant,
there were instances when communication between two offices broke
(especially in case of abroad offices). It was thought that it's better
to have DB server in each location in case of VPN tunnel failure;
frankly, I don't know if that's working.
Since Linux users have their homedirs on AFS, when AFS servers fail for
any reason, the only thing they can do is to reset their station and
enter emergency mode (ie. login with no own homedir, but allowing to
work), losing current work in progress. Again, I don't know if it'd be
enough to have fileserver
w/o DB server in given location. I guess I'll
go with switching abroad servers to "clone" mode (I didn't even know
something like this exists).
As for moving - I meant directory traversal, "cd". Although I checked
yesterday and it was fluent, so I'm not sure what was happening before.
As for relocating volumes - I'm afraid vos dump/restore is not an
option, because user can't work in the meantime. I asked about number
of cells because it seems normal to me that an organization has one
cell, but I read in the book (very old, I know) that it might be a
standard practice to create one or more cells per department.
Next question - we're only storing homedirs and groupshared on AFS,
There's no point to store software, local harddrives are way faster and
we have quite robust update system in place. So, no RO.
2/5 MBps on 1 Gbps. Now it doesn't sound that good, does it? :-)
xstat_fs_test - I'll try it and get back with results.
Thanks for all your answers :-)