In UCS - how to verify QoS queue usage

[mmedwid]

I am having some problems with VMWare connectivity to Netapp storage via the Cisco UCS.  Have found no errors at any ethernet interface and no errors on the Netapp side.  One thing I would like to eliminate is the possiblity of QoS shooting us in the foot.  In UCS how can I view what impact QoS is having on traffic?  What traffic has been throttled due to QoS policies?  What if any dropped?  Thank you!

[Robert Burns]

Before we jump into QoS as the culprit we need more information.

- Exact topology between UCS and the Netapp

- Explain "having osme problems with VMware connectivity".  Are you seeing poor performance, disconnects etc?

- Which blade, adaptor are you using?

- UCSM Fw version?

- VMware Driver version for the adapter (ethtool -i vmnicX)

From the equipment tab you can drill down to each vNIC and look at the stats on the right.  Here you can see any drops, errors, CRC, Pause framce (PPP) etc. 

Regards,

Robert

[mmedwid]

UCS-10Gbps fiber-->Nexus 5548--10Gbps fiber-->Netapp filer

VMware complains of latency and storage disconnection in Events viewer.  Guest VMs become unstable and fail to operate.  SNMP monitoring of the guests start to fail.  The guests no longer can be viewed through the console viewer in vSphere client.  The ESXi host becomes so unstable that only a cold reboot will bring it back.  And you can not vmotion off the servers, DRS fails utterly.

Blade PID N20-B6625-1, has Intel Xeon E549 CPU, adapter N20-AC0002

UCSM v2.0(1w)

vmnix is E1000

Looking at the stats.  Thank you very much.