Riaan,

I would venture a guess that the problem is throughput related. Even
though you have a GB connection the fact that you are streaming 1 flat
file becomes an issue. The client that is pushing the data to your
Networker Server can't push it fast enough. I have similar issues with
clients that are pushing several large flat files (over 20 GB each) and
the backup actually finishes but because the data rate drops to either
kilobytes and even bytes per second, Networker assumes the backup has
not completed. I think this is proven by the fact that backups made
using the BSM for SQL are fine. Multiple streams are pushed
simultaneously to Networker so the data rate is higher. You don't
mention what drive technology you are using but I would guess LTO. The
fastest backups I get are SQL. I push multiple streams to 1 LTO2 Tape
drive and I can easily hit the LTO2 write limit of 80 MB/sec. I am on GB
Ethernet which tops out at about 35 MB/sec (the Cat 5e spec is only
rated at 350-400 MHz which equates to 35-40 MB/sec with overhead the
rate is usually 33-35 MB /sec) but I get 3x1 compression so the speed
increases as it hits tape (I have a 64 MB cache on each tape drive).  I
have gotten 700 GB on an LTO2 tape with SQL backups.


Thanks,

Dana

-----Original Message-----
From: Legato NetWorker discussion [mailto:NETWORKER AT LISTSERV.TEMPLE DOT EDU]
On Behalf Of Riaan Louwrens
Sent: Friday, August 11, 2006 9:19 AM
To: NETWORKER AT LISTSERV.TEMPLE DOT EDU
Subject: Re: Dropped Connection Ports

Hi Howard, 

Thanks for the reply, I have taken this into account - so group
inactivity was already set to 0 (I haven't tried a large timeout yet
...)

I know the TCP parameter is for so called firewall timeouts - however,
the symptom I am having on the layer 3 switch is exactly the same -
packets get dropped that are inactive for an extended period. It seemed
like 2 hours - which I believe is the default setting for tcp/ip.

We have been told that the fact that the 2 machine are on the same
switch - but different vlan's should not have an affect - but we will be
testing it this weekend.

Any other input is appreciated!

Regards,
Riaan

_________________________
Channel Data Service Desk

Tel: +27 (0) 860 235 235
Fax: +27 (0) 11 235 7777
Servicedesk Email: servicedesk AT channeldata.co DOT za
Website: http://www.channeldata.co.za

The message is intended only for the use of the people to whom it is
addressed.
It may contain information that is privileged and confidential.
Accordingly any unauthorised use is strictly prohibited.
If you are not the intended recipient, please contact the sender as soon
as possible.


-----Original Message-----
From: Legato NetWorker discussion [mailto:NETWORKER AT LISTSERV.TEMPLE DOT EDU]
On Behalf Of Howard Martin
Sent: 11 August 2006 12:33
To: NETWORKER AT LISTSERV.TEMPLE DOT EDU
Subject: Re: [Networker] Dropped Connection Ports

Probably you just need to set the groups inactivity timeout value to
either 0 (never times out) or to a value larger than the worst case
backup time. On the GUI Customize - Groups - view hidden - Inactivity
timeout (value in minutes).
The tcp_keepalive_timer is only to stop firewall timeouts from messing
up the backups.

To sign off this list, send email to listserv AT listserv.temple DOT edu and
type "signoff networker" in the body of the email. Please write to
networker-request AT listserv.temple DOT edu if you have any problems wit this
list. You can access the archives at
http://listserv.temple.edu/archives/networker.html or via RSS at
http://listserv.temple.edu/cgi-bin/wa?RSS&L=NETWORKER

To sign off this list, send email to listserv AT listserv.temple DOT edu and 
type "signoff networker" in the
body of the email. Please write to networker-request AT listserv.temple DOT edu 
if you have any problems
wit this list. You can access the archives at 
http://listserv.temple.edu/archives/networker.html or
via RSS at http://listserv.temple.edu/cgi-bin/wa?RSS&L=NETWORKER