Question : Problem: EMC SAN Performance logging

My company has been having some issues with performance on a variety of our applications, and I wonder if it could be performance issues with our SAN.

We have logging enabled which views the following....

HBA Port errors
Device Errors
Resets
I/O Count
IOPS
BPS (Bits per second)

Well when logging the HBA Port errors, device errors, and resets remain zero, but the I/O count remains high.

How high is to high? How do I know the limit is being exceeded? Is there some kind of baseline or limit?

Answer : Problem: EMC SAN Performance logging

Thanks for that. The only thing out of the ordinary is that write cache is disabled because you have no Standby Power Supplies (SPS). The SPS are a mini UPS that keeps the array running for about 2 minutes after a power failure so that the array can flush the contents of write cache out to the vault area on the first 5 disks. No SPS = no write cache = pretty ordinary write performance.

I'd be very surprised if you purcahsed a CX300 without SPSs - it is more likely that they aren't connected correctly. Your CX300 contains a 3U cabinet chock full'o disks. Beneath that (typically) is a 1U tray with 1 or 2 SPS in it. From the rear of the cabinet, you should see the main power cables going from the supply to the SPS, then a shorter mains cable going from the SPS to the CX300. There is also a small signal cable that tells the array what state the SPS is in.

Can you check the cables please and post what you find? A photograph would be helpful too, if you can.
Random Solutions  
 
programming4us programming4us