Re: amanda over wireless network
2005-11-17 15:25:56
->>In response to your message<<-
--received from Paul Bijnens--
>
> Paul Yeatman wrote:
> >Hi, don't seem to find much concerning "wireless" when searching the
> >mailing list.
> >
> >I'm attempting an amanda backup of a computer on the wireless network
> >for the first time. Just preceeding this is compiling amanda on an
> >Apple powerpc and getting that to work. This has gone well. The
> >number of Apple laptops in our research group has been noticably
> >increasing as has the need for a way to back them up. I'd like to use
> >amanda for this if I can make it work. Adding to this is that several
> >such Apple laptop users have so far exclusively used the wireless
> >network. I'm just now experimenting with whether it will work to back
> >up these laptops over a wireless network using amanda.
>
> It should work, but I'm not sure what happens to the ether when 10
> laptops start sending each 10 gigabyte over the precious resource like
> the shared radiowaves all accessing the Access Point.
> Lot's of collisions I guess.
> I still remember when we did not have switch, only hubs (10Mb), and all
> ethernet packets where seen by all computers (only 20 or so!). Copying
> anything larger than 100 megabyte over the network made the rest of the
> network crawl like a snail.
>
>
> >I have successfully performed this feat on a home directory just over 3
> >gigs. When I try the full "/" partiion (17 gigs), however. it times
> >out during the estimate. I had already increased the estimate timeout
> >on the server side to be 900s. From where it appears to be failing,
> >given the amandad log on the client, increasing the timeout limit more
> >won't help
> >
> > .
> > .
> > .
> > Amanda 2.4 REP HANDLE 001-3003BD2A55000000 SEQ 1132177077
> > OPTIONS features=fffffeff9ffe7f;
> > / 0 SIZE 17648380
> > ----
> >
> > amandad: time 1249.536: dgram_recv: timeout after 10 seconds
>
> Seems that the estimate took 1249 seconds, and indeed, 349 seconds
> earlier, the server gave up waiting for this client.
> You should increase etimeout to at least 1300 (better double that,
> because next time, amanda could ask an estimate for level 0, level N,
> and level N+1 for each disklistentry!)
>
Ah, thank you so much! I didn't catch that! I saw only the
"timeout after 10 seconds"
which I couldn't see as having any relation to the 900s I was
allowing. I very much suppose that will do it as long as I don't run
into a problem with the your first concern mentioned.
Thanks again!
Paul
|
|
|