And here is the fourth. Most unexplainable performance problems are caused
by auto-negotiation in a 100TX environment (not Gigabit).
Any Admin that has been involved in a larger environment can back this up.

BUT there is something to be aware off. There is a very good reason why
not-using auto-negotiation fails in some environments. 
Most of the more modern, off-the-shelf, affordable low-cost switches (the
type that allows your accountant to save a lot of money) can not be
configured. If you follow our advise to set the client to 100 Full-Duplex
then backups will fail. Why? Because most, if not all, will fallback to 100
Half-Duplex if autonegotiation (can not be disabled on switch) fails. This
means you have to set the client to 100-Half or auto.

The best solution for a reliable backup is to used switches that can be
managed and set all ports to 100-FD.

Bye
Ernie

-----Original Message-----
From: Legato NetWorker discussion [mailto:NETWORKER AT LISTSERV.TEMPLE DOT EDU] 
On
Behalf Of Dag Nygren
Sent: Saturday, October 29, 2005 1:29 PM
To: NETWORKER AT LISTSERV.TEMPLE DOT EDU
Subject: Re: [Networker] Compare Netoworker to other backup products


> On Sat, 29 Oct 2005, Davina Treiber wrote:
> 
> DT> Oscar Olsson wrote:
> DT> > On Fri, 28 Oct 2005, Ty Young wrote:
> DT> > 
> DT> > TY> I would say 98% is quite good.
> DT> > TY> TY> In my experience the Windows clients talk very nicely 
> DT> > TY> TY> with Windows
> DT> > or UNIX
> DT> > TY> backup servers (NetWorker) and the failures I've seen with 
> DT> > TY> clients are usually due to
> DT> > TY> TY> (a) client NICs left at auto-negotiate instead of being 
> DT> > TY> TY> forced
> DT> > 100/full
> DT> > 
> DT> > Say what? Using anything else than auto-negotiation is stupid. 
> DT> > Auto negotiation has been a non-issue since the late 90s. 
> DT> > However, I have seen plenty of examples where a clueless 
> DT> > administrator has turned off auto-negotiation on the client, but 
> DT> > left auto negotiation on on the switch. This ofcourse results in 
> DT> > that the switch will regard the port as being in 100mbit/hdx 
> DT> > mode, which will result in plenty of late collisions and really 
> DT> > poor performance.
> DT> > 
> DT> > Please make yourself and the systems users a favour by using 
> DT> > autoneg everywhere.
> DT> I beg to differ. I find it much more reliable to force everything 
> DT> to 100/FD. Auto-neg settings are the number one cause of slow and 
> DT> intermttently failing backups in most of the environments I have 
> DT> worked in, especially for Windows clients.
> 
> OK, you are the second one who tells me this. But I still claim that I
> have seen several cases of poor performance in different brand 
> environments where speed and duplex has been set inconsistently. This 
> includes Cisco, HP and D-link switches.

And here is a third one!
I installed 10 -20 bigger Networker systems last year and saw about 50% of
these failing to negotiate decently under auto-negostiate. Usually they
ended up with one end 100/half and the other 100/full.

So IMHO it is really stupid to trust the auto-stuff. And most of the times
you dont't even nitice the poor performance until installing Networker...

And it is REALLY hard to convince the network-guys that Their Network has a
flaw in it ;-)

Dag

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