MDS Switches Provide and Enterprise Class SAN Fabric for Demanding Workloads

For raw throughput performance, we haven't experienced any issues directly relating to the [Cisco MDS 9000 Series Multilayer Switches], these issues have been up or downstream. We are not driving unified policies across the datacenter in our deployment, policy definition is done at a standards/governance layer and each environment is individually brought up to compliance. I would like to drive it further with automation, but having an applicable gap between updating A/B fabrics is great from a change management perspective. Automation drives efficiency but can add risk depending on implementation approach.

We tend to not deploy MDS's in an ISL deployment pattern, rather deploying a single unit with high enough port density to suit our requirements. Our standard switches are either 9248 or 9396's. My only issue is the high cost of adding additional port licenses "golden screwdrivers" for existing infrastructure. Historically its been cheaper to add a new switch and ISL than add additional licenses if you only have 1/4 to 1/2 licensed.

We have not experienced any specific critical or zero-day exploits that have required an urgent update of code. We leverage tools like Prime for AAA which meets our security requirements.

[Cisco MDS 9000 Series Multilayer Switches] provide an enterprise class SAN fabric for the consumption of storage devices across a multitude of devices within the datacenter. MDS is well suited to provide storage presentation to Virtualized environments (VMware in our case) and for bare metal hosts. The 9200 and 9300 family are capable to provide this without the requirement for a larger director switch within the datacenter.

Whilst configuration through the CLI is preferred, DCNM still falls short of being able to provide a single pane of glass without additional expensive licenses.