[Samba] (no subject)
joel.franco at gmail.com
Tue Nov 29 14:22:24 GMT 2005
| self-powered by
| Debian Linux
| : :' :
| `. `'
On Seg Nov 28 05 12:26, Gerald (Jerry) Carter wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> Joel Franco wrote:
> | Empirically, when a NT based is added to the
> | existing network, the database application runs
> | slowly (access data) in this client machine
> | and in others NT machines that are added to the LAN.
> | If i substitute this samba server with a Win2k3
> | Standard Server, the application latency backs
> | to the original (or close), before the NT clients.
> | This was a big deception to me and since then, i'm
> | trying to find why this happens but i'm specialist.
> What version of Samba are you running?
But this problem exists since old samba versions.
> | I have observed in the ethereal sniffer that exists a
> | lot of "Locking AndX Request" and "Locking AndX Response"
> | that is highly "ping pong" communication between
> | the server and the client, that certainly don't
> | permit a good brute transfer. I understand that must
> | exist a lock mechanism to not corrupt the file
> | database shared between others stations.
> Most people running pc based database apps disable
> oplocks on the server. Can you run a quick test
> with and without oplocks in the Samba share.
The oplocks warn that could be problem on this enabled. If a client
machine crash, the database could be corrupted, correct?
Should I understand the oplocks is a tradeoff and have to be choosed
between speed and corrupt probability?
> cheers, jerry
> Alleviating the pain of Windows(tm) ------- http://www.samba.org
> GnuPG Key ----- http://www.plainjoe.org/gpg_public.asc
> "There's an anonymous coward in all of us." --anonymous
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.0 (GNU/Linux)
> Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org
> -----END PGP SIGNATURE-----
More information about the samba