MSA1000 constant activity despite not being connected to anything

[Adrian Graham_1]

Folks,

Got a dual controller MSA1000 here running 7.2 firmware, active/active. 13 300gb drives in 2 LUNs. No parity init or any other activity, currently the single host it's connected to is powered off.

I know these will do background integrity checks whilst idle, but should the drive lights be constantly active with one controller being thrashed? Left controller looks like something's tasking it hard!

I can provide a show tech_support if needed, but it doesn't show anything of note, there's only 2 entries in the event log...

Cheers

[Sivaramakrishna_1]

Hi,

Can i know about which LED you are talking about on the HDD. Generally if the IO Operation are taking place then the one of the LED (activity LED) will be blinking.

There will be 3 LEDs on the HDD.

1. Activity status On = Drive activity
Flashing = High activity on the drive or drive is being configured as part of an array.
Off = No drive activity

2. Online status On = Drive is part of an array and is currently working.
Flashing = Drive is actively online.
Off = Drive is offline.

3. Fault status On = Drive failure
Flashing = Fault-process activity
Off = No fault-process activity.

The numbers from Top to Bottom.

[Adrian Graham_1]

Hi,

The activity light is almost constantly on and the left controller is running at 33% CPU useage with a '% Time Active Reqs' of 100% which I take to mean it's constantly being polled or accessed. I've attached a picture showing the lights and left controller.

Structure version: 101
% CPU utilization: 33%
Command count: 0
Avg Command Latency: 0.00ms
% Time Active Reqs: 100.0%
Max Outst Cmd Lists: 0
Logical Req Count: 0
Avg Free Phys Reqs: 0
Avg Free Log. Reqs: 0
Avg Free Xbuf Sects: 0
Avg Lrgst Avail Xb S: 0
Avg Free CPU DRAM KB: 0
Max DMA Xfer Q Depth: 0
% Time DMA Xfr Stall: 0.0%
Avg Locked Stripes: 0
Avg Stripes Waiting: 0
Avg Completions/Sec: 0.0
Avg Log Req Comp/Sec: 0.0

[Sivaramakrishna_1]

Hmm..

It seems to be strange behaviour. Below are the observations as per the LED status.

LED 3, 4 on: Indicating that requests are outstanding on both scsi bus 0 and 1.

LED 5 is off: stats that Logical IO is stopped.

But the LED 8 is on that indicating Controller is Idle.(but your data is showing CPU 33%)

LED 9, 10, 11 is ON indicating controller is idle.

[Adrian Graham_1]

It's weird, I've never seen behaviour like it and we've installed loads of MSA1000/1500s. We can change controllers and it still happens. Destroy the units and activity stops, remake them and as soon as the parity init has finished it's off again!

I found that one of the drives had 6000 errors whilst the rest had 30-80 so wondered if that drive was upsetting the bus, but it's now out and we still see oddness.