I have a customer with a VNX 5300 with 2 200GB SD for FAST cache. I’ve done the SP/NAR collect on the array and presented to the customer already.
Question is is there a way to know HOW the Fast Cache SSD’s are being used by the VNX? If the existing capacity is underutilized, over utilized, etc, or additional capacity is needed. Are there other tools available to get that info ?
My understanding is that since we cannot specifically allocated sections on the fast cache, but just enable them on the pools we don’t control that area right?
I have also looked at the SP/NAR the Fast Cache does look being utilized well. You can add 2 more SSD’s ( to max it out since 4 is the max on a VNX5300). I’ve also seen that they have 5 RAID groups and have FAST Cache enabled on only one of them.
Beyond that, I am not sure what other info I can get. Any help is appreciated. What other options do I have?
Thanks in advance for your help.
I would not try to find out how much of the FAST Cache is actually being used; always assume it is pretty much full while enabled (after it has had time to 'warm up').
Instead, try to analyze the read/write hit rate, dirty pages percentage, and how much throughput is on the individual EFDs.
I find Unisphere Analyzer to be quite useful when analyzing FAST Cache performance. This can help:
Hope this helps,
Unified Performance Support
You can try to implement FAST VP and FAST Cache will be much faster to work with the system. Use FAST VP within storage pools.
EMC internal TC's and partners with USPEED access have tools that can give visibility inside the Fast Cache at the drive, RG and private LUN levels. This will show if FC is saturated. Do you have a resource at EMC that you could reach out to?
I agree, contact EMC reps and they can help with analysis for the FastCache performance.. and also looking at the skew of the workloads, with ktrace and such.