Re: gb issue

Spanning-tree is your solution, but since you have unmanaged switches (or at
least one) that won't work. You have loop in your network and that's whets
"breaking" Your only fix with your current hardware is to leave that
connection unplugged, or unplug another link. Basically you can not have a
complete loop. Trace your finger on your drawing, you should never be able
to complete a full circle

http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/sw_ntman/cwsimain/cwsi2/cwsiug2/vlan2/stpapp.htm


"Terri S" <TS@xxxxxxxx> wrote in message
news:OZXQl3L1FHA.268@xxxxxxxxxxxxxxxxxxxxxxx
> I have the cable guys installed two CAT6 cables in between the two rooms,
> each room I have a All-GB-port switch (unmanaged, shown on the drawing,
one
> 16 port and one 8 ports)
>
> in closet B is where we have the SAN and are planning on putting most of
our
> servers, I decided to build a GB backbone before all this happen, I
already
> had about 60-70 ports of workstations occupied in room B (SW24-ABC), in
> which they are on 10/100 ports (all unmanaged switches as well) with two
GB
> uplink port (on the drawing I showed the way I chained them) so they can
all
> talk to room A (from SW24-ABC to SW8) where all the servers were at that
> time
>
>
> We have some servers now connected to SW16, including the SAN file server
> which everybody is going to hit once we open the door to the SAN storage,
> the cable marked " 1 " is the cable it rely on to talk to the rest of the
> network (room A has about 60-70 percent of workstations and half of the
> servers still there), that means (correct me if I am wrong) before people
> can get to the SAN, it has to go through Switch ABC (SW24-ABC s on the
> drawing), the GB connection has to be divided by the switch chains,
>
>
>
> With that in mind, I activated the other original CAT6 run (which runs
> between the rooms, marked "2" on the drawing") by connecting switches SW16
> and SW8 in hope that people can get faster access to the SAN and the
servers
> on SW16, but I immediately get people's calls that network is not working
> (especially the email and the internet) and it back to work if I
disconnect
> the cable marked "2"
>
>
>
> so does means this setting does not allow 2 uplink cables on SW16 ? should
I
> have disconnect 1 before connect 2 ?
>
>
>
> note that I put down on the drawing that the SW8 also goes to other 10/100
> switches in room A, which majority of user there complained about the down
> connection, I also have all the users complain about email not work
because
> the exchange server is connected to the SW16, I didn't have time to check
> around but I believe the outage was across the broad as soon as I
connected
> cable 2
>
>
>
>
>


.

Relevant Pages

  • Re: when to nas or san
    ... >>>Anyone have any good pointers to documents on when to use Nas or a San ... >running a multitude of servers with exchange2k, sql server, a soon to be ... Enterprise class switches are still McData and now Cisco. ...
    (comp.arch.storage)
  • Re: GB Switch Connection
    ... switches which support spanning tree protocol. ... A gige link should be more htan enough for 60 clients hitting a file server. ... > We have some servers now connected to SW16, including the SAN file server ... > drawing), the GB connection has to be divided by the switch chains, ...
    (microsoft.public.windows.server.setup)
  • Re: Overall networking issue...
    ... At the moment I am thinking about buying a new gigabit backbone switch that I can connect all of the other switches to, and this gigabit switch should have enough extra ports so that all of the internal servers can connect here directly. ... This should give a more acceptable speed for the clients connection to the fileserver. ...
    (comp.os.linux.networking)
  • Re: GB Switch Issues
    ... >I have the cable guys installed two CAT6 cables in between the two rooms, ... > our servers, I decided to build a GB backbone before all this happen, I ... > We have some servers now connected to SW16, including the SAN file server ... > drawing), the GB connection has to be divided by the switch chains, ...
    (microsoft.public.windows.server.setup)
  • Cluster goes down the moment Uplink is connected
    ... I have two Linux AS5 servers running in cluster mode with two HP procurve ... switches and MSA 1500 SAN, both the servers have one on-board LAN card.The ... The SAN storage gets mounted on any of the ...
    (RedHat)
Loading