Its a switch. The hosts are all off vlans that are trunked to the switch from the outside network. Either you have one access vlan for all the blades, or if it supports it, you have trunks and you can then assign the blades to one of the trunked vlans. The switch itself needs a default gateway so you can manage it, but that has nothing to do with connectivity to the blades, since that is all layer 2. The outside network knows how to get to the blades by arping for them in the right vlan, and the blades know how to get to the outside world via an IP and default gateway you set on each blade.
Not saying the requirement is not legit, but why do you need a static route to a blade/host, and how would that get to out to your network when there is no routing relationship (ospf, rip, eigrp)? Blade centers should be all layer 2.
Its a temporary solution for a wireless access point, as two people need to get access to the switch. I wanted to add a *specifc* host route so the switch knows to send these packets and a default route for the other guys.
So put the static on the router that owns the vlan where the blades are. I think that should provide the same functionality. A route on the switch wouldn't do any good since first, it would only impact mgmt traffic to the switch and not the blades themselves (the blades send traffic to the gateway which is defined on the blade which is completely separate from the switch's default gateway, other than they probably happen to be the same thing). You need to place the route on the actual default gateway router.
On another note, best practice with an IBM blade center is to not IP the switch and mgmt interfaces in the same vlan as the blades. And technically, the switch's IP is accessed via the physical connection to the management port, not through the connections to the switch which are only used for l2 access to the blades. Therefore, you would have two ports in vlan A going to the mgmt ports, and two or more ports in VLAN B going to the switch. However, the 4 IPs on the blade center (2 mgmt and 2 switch) would all be in vlan A. Vlan B would only have IPs of the blades themselves. If you do not do this, and instead have all ports/IPs in the same vlan, traffic will loop within the blade center as the blades send out traffic and the external switch will send it right back into the blade center on the mgmt ports. IBM has known this for quite sometime, and drops the traffic, but you will see logs entries for this behavior.
Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.