Yeah wanted to see if the standard NIC's could be configured for fail
over and for increased throughput. The whole ISP is Solaris! 5 different
countries with around 40 servers in each.........
I'm a networking guy so they don't like me setting up their solaris
boxes..........So mirroring notes not much use to me! cheers for the nfo
though.
Gossip: We just had an E10k stolen..................The insurers are
going wild.................
At least Foundry have a sense of humour... not only is their CLI almost
totally Cisco compliant you get things like this too........
Cisco HSRP
Foundry FSRP
Cisco are gonna buy them up..............
Nice
L8r
-----Original Message-----
From: Sach Jobb [mailto:[EMAIL PROTECTED]]
Sent: 16 July 2001 03:00
To: Mike Flynn
Cc: [EMAIL PROTECTED]
Subject: RE: Solaris ahhhhhhhhhhhhhhh
Yes, i LOVE Foundry.
I have been working with their switches for some four years now and they
are by far the best switches in terms of supporting a variety of
protocols
->AND<- having great throughput. Just thinking about them makes me
happy.
It's interesting to see that they are branching out a bit from just
switches. Let me know how the ServerIron works out.
Is your entire ISP Solaris based? Interesting. So this card is just to
provide NIC failover? Or increased bandwidth? Or both?
On a similiar topic i was just setting up disk mirroring (RAID 1) on a
250
and a 450 for one of our clients using Solaris's built in DiskSuite and
i'm REALLY impressed. It's very simple and elegant. If you are
interested
i'll mail out my notes when i'm done.
Cheers,
sach
On Mon, 16 Jul 2001, Mike Flynn wrote:
> The solaris boxes are doing everything an ISP needs. Web, DNS , POP,
> SMTP etc...........
>
> Thorsten pointed me to a good network card
> http://www.znyx.com/drivers/os/solaris.htm that does the job. Just
> thought you would be able to do it neatly with the built in
> cards..............
>
> We are going to use Foundry ServerIron's which seem way better than
the
> Cisco director..................
>
> Thx f da help.............
>
>
> -----Original Message-----
> From: Sach Jobb [mailto:[EMAIL PROTECTED]]
> Sent: 13 July 2001 15:23
> To: Thorsten Sideb0ard
> Cc: Mike Flynn; [EMAIL PROTECTED]
> Subject: Re: Solaris ahhhhhhhhhhhhhhh
>
> I guess i depends on what the servers are being used for but, don't
use
> usually use some sort of front-ending sort of box (ala cisco
> localdirector) to handle this sort of thing.
>
> What are the solaris boxes doing?
>
> sach
>
>
> On Thu, 12 Jul 2001, Thorsten Sideb0ard wrote:
>
> >
> > Sitting in a little Internet cafe i Barcelona at the moment,
> > so i can't do too much research.
> >
> > Mostly when i've read up on load balancing on the servers,
> > all the solutions have been slightly dodgy.
> >
> > i think you have three options,
> > a simple script that checks the state of each interface and changes
> route
> > just using ifconfig would be your lo-fi method, probably not what
your
> > looking for.
> >
> > A routing protocol, probably gated. Don't know too much about using
> this,
> > but from what i've read i think it could do the job.
> >
> > and i believe there are various hardware solutions.
> > One that looks good is
> > http://www.znyx.com/drivers/os/solaris.htm
> > provides "
> > link aggregation (for greater overall network throughput)
> > fast failover (automatic link failure detection/correction)
> > automatic load balancing (improve network efficiency) "
> >
> >
> > Is this for web servers at all?
> > There are a few things aimed specifically at load balancing web
sites.
> > Oh, my time is running out, gotta dash.
> >
> > Mail me off list with your current phone number and i'll give ya a
> call as
> > well, tell you the story behind the barcelona trip!
> >
> > later,
> > thor
> >
> >
> >
> >
> > On Fri, 13 Jul 2001, Mike Flynn wrote:
> >
> > > Thor
> > >
> > > How ya doing man? Long time no hear. Got a new job at ISP running
> > > network..... The server guys
> > > Need some help buddie.... we got some solaris issue with
connecting
> > > solaris 2.6,2.7, or 2.8 to 2 different catalysts.
> > >
> > > We want redundancy and would like both links to the different
cat's
> live
> > > at the same time and to provide fail over in the event of either
cat
> > > loss or link loss. The Sun's just seem to use one link by default
> and
> > > won't use secondary in the event of failure......
> > >
> > > Have changed the local_mac_address in the EEPROM to get unique
MAC's
> on
> > > different interfaces.........
> > >
> > > Anyone done this?????????????????????????????????????????
> > >
> >
> >
>
>
--
/*
sach jobb
sugoi consulting
1177 polk street
san francisco, ca 94109
415.345.8872 (w)
415.345.8871 (f)
415.559.5483 (c)
[EMAIL PROTECTED]
%s/windows/linux/g
*/
