Are you referring to CPU utilization or memory? I think, the 70% SP Utilization needs to be investigated and verified - considering you have empty enclosures. It is very much related to your configuration and usage - I suggest to engage EMC Support or your local EMC Team to help checking this out.

Thanks,
Sandip

Hypothetically speaking - you need to check why the SP is highly utilized - it's not common that partially populated CX3-80 will have high SP utilization. Also you need to check the total IOPS and Bandwidth for all the LUNs. I would have analyzed the configuration and utilization more before adding further disks to the array if single SP is having 75% utilization. It's not fare to say that the array is full - you can definitely add more disks - but be careful and understand the situation.

On this reference, I 'll suggest to take a look at the following two docs on powerlink -

"EMC CLARiiON Fibre Channel Storage Fundamentals - Technology Concepts and Business Considerations "
http://powerlink.emc.com/km/live1/en_US/Offering_Technical/White_Paper/H1049_emc_clariion_fibre_channel_storage_fundamentals_ldv.pdf

"EMC CLARiiON Best Practices for Fibre Channel Storage: FLARE Release 26 Firmware Update - Best Practices Planning"
http://powerlink.emc.com/km/live1/en_US/Offering_Technical/White_Paper/H2358_clariion_best_prac_fibre_chnl_wp_ldv.pd

Also - please take a look at the following old thread in this forum -
http://forums.emc.com/forums/thread.jspa?messageID=586473򏋩

And anytime - you may contact EMC Support who can provide you more help and guidelines if needed.
Thanks,
Sandip

To clarify, this is a theoretical question. I'm just trying to determine if you can just fill up a Clariion with the maximum number of drives regardless of performance or whether you need to consider the existing workload.

For example, say you have only added 240 of the possible 480 drives to CX3-80 and analyser shows %utilisation for each SP running at %75 throughout the day. Should that Clariion be considered full and no further disks added or would it be ok to keep adding the other 240 disks?

I believe that 70% is basically a watermark for performance. Once you go past it you risk possible issues, like an impact on disk rebuilds.

I also think I remember hearing from one of the training classes that you do not want to set SP utilization higher than 80%. I hope this helps.

Thanks MPaez and Sandip for very helpful replies to my question.

One last point to consider, if both SP's are constantly at 50+ % and you have a SP failure, there will a massive load put on the remaining SP (eg 100%).

We have a cx3-80 with only 180 spindle (300 15k FC). We are seeing SP average around 40%. (200+ vm farm). We are unlikely to add any more spindles to this array.

A clarification - the loading on the Storage Processors is not fully independent; that is, there are operations that are duplicated. For example, when both Storage Processors are running, they are both monitoring the status of all the objects (disks, RAID Groups, LUNs, etc.) within the array. Should one Storage Processor fail, that operational load does not transfer to its peer, as the peer is already performing that task. The take-away is that CPU Utilization is not simply/directly additive between the two Storage Processors. Like most peformance items, there isn't a single answer that applies to all configurations/uses. We are aiming to include guidance on this area in an update to the Best Practices document; as Sandip noted above you may want to reach out to local EMC personnel if you have a question on your current configuration.

Regards,
DGM