is this a bad design for edge switches

[baselzind]

my building floors edge switches are connected in cascading (as attached) with the first switch connected to core 1 and the last switch connected to core2. is this a bad design? would it cause loops?

[marce1000]

 - If possible connect all edge switches to the core preferably in  redundancy mode (link to each core component) and core-redundancy too, otherwise you are loading the other edges with traffic intended for the core.

 M.

[Jitendra Kumar]

not bad !! but the design should be redundancy each switch,s should be connected with a two-way. If one cable is faulty or the uplink port has an issue then, job should be taken care of by standby connectivity.

[MHM Cisco World]

the STP will BLK one link,
if the Core1 and Core2 is interconnect and use high BW link then one of the Access SW link will be BLK 
if the Core1 and Core2 is interconnect and use same BW as Access SW interconnect Link then there is chance that Core Interconnect is BLK and this BAD design.
cisco always recommend triangle design if you could not acheive that, then 
make Access SW interconnect to each other via PO <this make if one link is down there is redenducy>
make the Access SW interconnect Link high cost than Core interconnect link.

[Joseph W. Doherty]

The floor switches are not "cascading" as in a logical one device stack, correct?

"is this a bad design?"

"Bad" not so much as perhaps less that "ideal", but often we're stuck with real world topology that's less that "ideal".

"would it cause loops?"

Cannot say, as you didn't describe how, or even if, core 1 and 2 are interconnected at L2 for those switches.  If there is a L2 connection, for those switches between the cores, you'll have an L2 loop, which is a very bad thing, although likely you're running some STP variant to logically break the physical loop.  (BTW, there are other methods to manage L2 physical loops, e.g. REP or FlexLink.)

Assuming you do have a physical loop, broken by a variant of STP, there are considerations, or optimal design, like what link to block, root placement, gateway placement and FHRP.

[baselzind]

well these switches aren't stacked so each switch have its own mgmt ip also core 1 and core 2 are layer 3 switches for intervlan routing and there is a cable connection between core 1 and core 2 as a trunk

[Joseph W. Doherty]

Alright, in that case, you're "stuck" with a less than optimal design, but even so, an "optimal" design might not be so much better it would make any difference.

What you can do, as I mentioned in my prior posting, is consider other issues when using such a topology.  Some of those might be significant.

I believe Cisco validated design documentation would cover such a use case.

[balaji.bandi]

Most cases we follow cisco CVD to deploy to best output, some cases we may lead to business needs and requirement, at the time we may need to understand the risk of deploying and known issues.

1. why not connect all the switches to core - what is the limitation

2. why do you need to consider this design (what challenges you have ?)

3. if they are Layer3 you do not have any issue with certain level

Looking forward your side inputs.

[baselzind]

limitation is that i only have two fiber uplinks from each floor so i cant connect all switches to the core

[MHM Cisco World]

and we not say that you couldnot, 
we say that you need to make the Cost of cascade SW interconnect link have cost that not BLK the Core interconnect trunk link.
that it.
also I want to mention, when you dont add the cascade SW , each SW have one GW which is SVI in of VLAN in Core, here the new SW have VLAN that found in both Cores ?

[balaji.bandi]

Then you can deploy

Core to other switch dual Link (I take this as stack ?)

From other switches to the rest of the dual links should give high availability.