[Samba] Copy to 3.0.9-2.3-SUSE dies after 1.2 gb copied

[Tom Peters]

I have Samba 3.0.9 on SuSE 9.2 on a P-III-500 with only 352mb of RAM. The 
share I'm copying to is on a SCSI RAID array handled by a Compaq Smart-2/P 
RAID Controller (rev 04) as identified by lspci. The RAID is all LVM and 
the share is ReiserFS-- maybe not the best choice for something which is so 
slow to write to.

Anyhow, the problem is that when I copy to the share "movies" it will do 
about 1.2 to 1.4 gb and die-- my windows XP client (either one of them) 
reports "The  specified network name is no longer available" and the only 
choice is to click OK.

If I use a drag'n'drop copy, I'm all done. If I use a slightly smarter 
tool, one that apparently retries, I click ok and the copy takes off again, 
apparently none the worse. If I give up the first time it fails, the files 
it did manage to copy were between 133,122kb and 344,332kb in size.

I was advised to try some smb.conf changes. In the global section, I added 
"client use spnego = no". In the section for the "movies" share, I added 
"use sendfile = no". Niether change seemed to make any difference.

I also added "write cache size = 262144" in that share. Also little help 
that I can tell-- maybe a little better.

I'm slightly confused that I get two log files for the same machine, both 
seemingly updated concurrently, one called samba-log.172.20.0.186 and one 
called samba-log.elrond

The one named for the machine name doesn't have any errors or warnings-- 
just file opens and closes. The one with the IP address has errors.This 
fragment is pretty representative:
[2005/03/09 13:52:04, 1] lib/util_sock.c:get_peer_name(1095)
   Gethostbyaddr failed for 172.20.0.186
[2005/03/09 13:52:04, 2] lib/access.c:check_access(324)
   Allowed connection from 172.20.0.186 (172.20.0.186)
[2005/03/09 13:52:04, 2] smbd/reply.c:reply_special(235)
   netbios connect: name1=TOLKIEN         name2=ELROND
[2005/03/09 13:52:04, 2] smbd/reply.c:reply_special(242)
   netbios connect: local=tolkien remote=elrond, name type = 0
[2005/03/09 14:23:12, 1] lib/util_sock.c:get_peer_name(1095)
   Gethostbyaddr failed for 172.20.0.186
[2005/03/09 14:23:12, 2] lib/access.c:check_access(324)
   Allowed connection from 172.20.0.186 (172.20.0.186)
[2005/03/09 14:23:12, 2] smbd/reply.c:reply_special(235)
   netbios connect: name1=TOLKIEN         name2=ELROND
[2005/03/09 14:23:12, 2] smbd/reply.c:reply_special(242)
   netbios connect: local=tolkien remote=elrond, name type = 0
[2005/03/09 14:56:38, 1] lib/util_sock.c:get_peer_name(1095)
   Gethostbyaddr failed for 172.20.0.186

I also have stuff in my messages log, specifically, I see this every 15 
minutes:

Mar  9 16:21:08 tolkien nmbd[5713]: [2005/03/09 16:21:08, 0] 
nmbd/nmbd_browsesync.c:find_domain_master_name_query_fail(353)
Mar  9 16:21:08 tolkien nmbd[5713]:   find_domain_master_name_query_fail:
Mar  9 16:21:08 tolkien nmbd[5713]:   Unable to find the Domain Master 
Browser name RIVENDELL<1b> for the workgroup RIVENDELL.
Mar  9 16:21:08 tolkien nmbd[5713]:   Unable to sync browse lists in this 
workgroup.

Any clue what tuning I need to be doing here? I'm pretty new to Linux and 
Samba and need a clue.

TIA,
Tom



[Praise] The only way to escape the personal corruption of praise is
to go on working. --Albert Einstein
--... ...--  -.. .  -. ----. --.- --.- -...
tpeters at nospam.mixcom.com   (remove "nospam") N9QQB (amateur radio)
"HEY YOU" (loud shouting)  WEB ADDRESS http//www.mixweb.com/tpeters
43° 7' 17.2" N by 88° 6' 28.9" W,  Elevation 815',  Grid Square EN53wc
WAN/LAN/Telcom Analyst, Tech Writer, MCP, Cisco Certified CCNA