Disk Speed Bandwidth Capped on SAS SSD drives

[mytime34]

To All,

 

I am running 4ea 7.68tb Enterprise SAS 12gb SSD drives and when I run Disk speed, I get the following message.

Bandwidth was capped on the folloiwng drives: Cache (sdm), Cache 2 (sdn), test (sdo), test2 (sdp)

 

2 of these drives are hitting 700mbps, but the other 2 are only hitting 350mbps. Is there a way to uncap the bandwidth?

 

These drives are capable of 900-1100mbps each, and has been tested on a windows machine.

 

Any thought or fixes would be great.

 

Thanks

[ChatNoir]

Please share your diagnostics on your next post.

[mytime34]

pughhome-diagnostics-20211118-0649.zip

[ChatNoir]

I am not great at deciphering the logs, but that does not seem OK.

 

Nov 18 00:16:59 Pughhome kernel: ACPI Warning: SystemIO range 0x0000000000000B00-0x0000000000000B08 conflicts with OpRegion 0x0000000000000B00-0x0000000000000B0F (\GSA1.SMBI) (20200925/utaddress-204)
Nov 18 00:16:59 Pughhome kernel: ACPI: If an ACPI driver is available for this device, you should use it instead of the native driver
Nov 18 00:16:59 Pughhome kernel: sfc 0000:08:00.0 (unnamed net_device) (uninitialized): Solarflare NIC detected
Nov 18 00:16:59 Pughhome kernel: ccp 0000:09:00.2: ccp enabled
Nov 18 00:16:59 Pughhome kernel: ccp 0000:48:00.2: enabling device (0000 -> 0002)
Nov 18 00:16:59 Pughhome kernel: ccp 0000:48:00.2: ccp enabled
Nov 18 00:16:59 Pughhome kernel: sfc 0000:08:00.0 (unnamed net_device) (uninitialized): Part Number : SFN7x02F
Nov 18 00:16:59 Pughhome kernel: sfc 0000:08:00.0: enabling device (0000 -> 0003)
Nov 18 00:16:59 Pughhome kernel: sfc 0000:08:00.0 (unnamed net_device) (uninitialized): no PTP support
Nov 18 00:16:59 Pughhome kernel: sfc 0000:08:00.0 (unnamed net_device) (uninitialized): Insufficient resources for 32 XDP event queues (12 other channels, max 32)
Nov 18 00:16:59 Pughhome kernel: random: crng init done
Nov 18 00:16:59 Pughhome kernel: random: 7 urandom warning(s) missed due to ratelimiting
Nov 18 00:16:59 Pughhome kernel: sfc 0000:08:00.1 (unnamed net_device) (uninitialized): Solarflare NIC detected
Nov 18 00:16:59 Pughhome kernel: mpt3sas version 35.100.00.00 loaded
Nov 18 00:16:59 Pughhome kernel: mpt3sas_cm0: 63 BIT PCI BUS DMA ADDRESSING SUPPORTED, total mem (65808864 kB)
Nov 18 00:16:59 Pughhome kernel: sfc 0000:08:00.1 (unnamed net_device) (uninitialized): Part Number : SFN7x02F
Nov 18 00:16:59 Pughhome kernel: sfc 0000:08:00.1: enabling device (0000 -> 0003)
Nov 18 00:16:59 Pughhome kernel: sfc 0000:08:00.1 (unnamed net_device) (uninitialized): no PTP support
Nov 18 00:16:59 Pughhome kernel: sfc 0000:08:00.1 (unnamed net_device) (uninitialized): Insufficient resources for 32 XDP event queues (12 other channels, max 32)

 

You might also have an old FW on your SAS HBA

Nov 18 00:16:59 Pughhome kernel: mpt3sas_cm0: LSISAS3008: FWVersion(15.00.02.00), ChipRevision(0x02), BiosVersion(08.35.04.00)

 

We'll see what the pros have to say.  

[JorgeB]

14 hours ago, mytime34 said:

Bandwidth was capped on the folloiwng drives:

Assuming this is during the simultaneous device test it suggests a controller/enclosure bottleneck, do they perform normally on the individual tests?

 

In doubt post a screenshot of the results.

[mytime34]

I have tried running the tests individually and simultaneous with the same results. Bandwidth is always capped on these 2 drives.

My 2 other sas ssd drives do not cap out, they run without issue.

[JorgeB]

That then suggests a device issue, they are different brands.