| 1 |
Bryan, |
| 2 |
OK, according to ethereal I'm only getting NFS V2 across the bus. |
| 3 |
Here's the way my two machines are set up: |
| 4 |
|
| 5 |
The new 250GB storage machine - 2.6.12-gentoo-r6 |
| 6 |
|
| 7 |
<*> NFS file system support |
| 8 |
[ ] Provide NFSv3 client support |
| 9 |
[ ] Provide NFSv4 client support (EXPERIMENTAL) |
| 10 |
[ ] Allow direct I/O on NFS files (EXPERIMENTAL) |
| 11 |
<*> NFS server support |
| 12 |
[*] Provide NFSv3 server support |
| 13 |
[ ] Provide NFSv4 server support (EXPERIMENTAL) |
| 14 |
[*] Provide NFS server over TCP support |
| 15 |
|
| 16 |
The mythbackend machine which is the NFS client: |
| 17 |
|
| 18 |
<*> NFS file system support |
| 19 |
[*] Provide NFSv3 client support |
| 20 |
[ ] Provide NFSv4 client support (EXPERIMENTAL) |
| 21 |
[ ] Allow direct I/O on NFS files (EXPERIMENTAL) |
| 22 |
<M> NFS server support |
| 23 |
[*] Provide NFSv3 server support |
| 24 |
[ ] Provide NFSv4 server support (EXPERIMENTAL) |
| 25 |
[*] Provide NFS server over TCP support |
| 26 |
|
| 27 |
I guess V2 support makes sense as that kernel only has V2 compiled in. |
| 28 |
Shall I jump to V3 or V4? |
| 29 |
|
| 30 |
thanks, |
| 31 |
Mark |
| 32 |
|
| 33 |
On 8/2/05, Bryan Whitehead <driver@×××××××××.net> wrote: |
| 34 |
> BTW, it could be you are using NFS v2 which is ONLY 32bit so you have the |
| 35 |
> 4gb filesize limit. |
| 36 |
> |
| 37 |
> run "nftstat -s" (on the server) and "nfsstat -c" (on the client) to see |
| 38 |
> what version of NFS you are using (note: what version of NFS you are using |
| 39 |
> is not related to the transport - udp/tcp). |
| 40 |
> |
| 41 |
> I use bigger than 4GB files on Linux server/client all the time to move |
| 42 |
> DVD iso's to machines with better burners... |
| 43 |
> |
| 44 |
> you are running the 2.6 kernel? |
| 45 |
> |
| 46 |
> On Tue, 2 Aug 2005, Bryan Whitehead wrote: |
| 47 |
> |
| 48 |
> > What filesystem are you exporting over NFS? |
| 49 |
> > |
| 50 |
> > On Tue, 2 Aug 2005, Mark Knecht wrote: |
| 51 |
> > |
| 52 |
> >> On 8/2/05, Matthew Cline <matthew.cline@×××××.com> wrote: |
| 53 |
> >> > On 8/2/05, Mark Knecht <markknecht@×××××.com> wrote: |
| 54 |
> >> > > |
| 55 |
> >> > > but how do I know it's being used? And how do I know that the rsize |
| 56 |
> >> > > option is being used? |
| 57 |
> >> > > |
| 58 |
> >> > > Thanks, |
| 59 |
> >> > > Mark |
| 60 |
> >> > |
| 61 |
> >> > Could you watch the traffic between the two using something like |
| 62 |
> >> > ethereal? This should tell you which protocol is being used. |
| 63 |
> >> > |
| 64 |
> >> |
| 65 |
> >> Hi Matt, |
| 66 |
> >> OK, ethereal was pretty easy to use, and it does indeed show that |
| 67 |
> >> I'm using TCP for packat transfer. I see a proto=NFS packet followed |
| 68 |
> >> by a number of TCP packets with sizes of 8K bytes so this seems to |
| 69 |
> >> verify that both options I was looking for ar indeed working. |
| 70 |
> >> |
| 71 |
> >> Thanks! |
| 72 |
> >> |
| 73 |
> >> Unfortunately this means I'm no closer to the root cause of my real |
| 74 |
> >> problem which is mythbackend shutting down without warning. It |
| 75 |
> >> happened again just a few minutes ago. This all started happening |
| 76 |
> >> after I brought this NFS mount on-line as storage for the mythbackend |
| 77 |
> >> server. I suppose I'll have to go back to the reduced storage option |
| 78 |
> >> (15 hours instead of 120 hours) and make sure that it's really this |
| 79 |
> >> disk/PC/network connection. |
| 80 |
> >> |
| 81 |
> >> Thanks again for your help. |
| 82 |
> >> |
| 83 |
> >> Cheers, |
| 84 |
> >> Mark |
| 85 |
> >> |
| 86 |
> >> |
| 87 |
> > |
| 88 |
> > |
| 89 |
> |
| 90 |
> -- |
| 91 |
> Bryan Whitehead |
| 92 |
> Email:driver@×××××××××.net |
| 93 |
> -- |
| 94 |
> gentoo-user@g.o mailing list |
| 95 |
> |
| 96 |
> |
| 97 |
|
| 98 |
-- |
| 99 |
gentoo-user@g.o mailing list |
Archives