Customer has two AFF8080EX.
The basic usage is as follows.
CPU AVG(%) | NFS IOPS | NetIn(KB/s) | NetOut(KB/s) | SSD Read(KB/s) | SSD write(KB/s) | |
AFF1 For Internet Virtualization Ontap 9.0, 960G x 24 | 28% | 24,919 | 180,991 | 68,046 | 112,120 | 110,870 |
AFF2 For Internal Office Ontap 9.2p1, 3.8TB x 42 | 69% | 30,508 | 263,343 | 219,176 | 233,427 | 131,624 |
This is for AFF2 sysstat -x
This is for AFF2 sysstatM
This is for snapshot background for AFF2
Cache Hit ratio
AFF2(For Internal Office)
workload_volume:SCWOS02-wid9819:ops:3729/s
workload_volume:SCWOS02-wid9819:read_io_type.cache:55%
workload_volume:VDISOOS05-wid9955:ops:4194/s
workload_volume:VDISOOS05-wid9955:read_io_type.cache:61%
workload_volume:VDISOOS06-wid23596:ops:5017/s
workload_volume:VDISOOS06-wid23596:read_io_type.cache:63%
workload_volume:VDISOOS07-wid28852:ops:10488/s
workload_volume:VDISOOS07-wid28852:read_io_type.cache:72%
AFF1(For Internet Virtualization)
workload_volume:VDIIOS02-wid2746:ops:6913/s
workload_volume:VDIIOS02-wid2746:read_io_type.cache:95%
workload_volume:VDIIOS04-wid23117:ops:8795/s
workload_volume:VDIIOS04-wid23117:read_io_type.cache:95%
workload_volume:VDIIMM02-wid28829:ops:16702/s
workload_volume:VDIIMM02-wid28829:read_io_type.cache:99%
And AFF2 storage performs snapmirror on six volumes at the same time with two target storage, and proceeds in 10-minute increments.
And AFF2 storage is running snapshot for six volumes per 5 minutes
And AFF2 storage is using Inline dedup/compression/compaction
I understand that CPU usage is high, but CPU usage seems to be too high compared to IO.
I doubt it is a normal situation.
The storage for the internal office is very important and can not be used in this high CPU state.
If this is a problem in their environment, they will go to the AFF A700.
If it is a problem of snapmirror or snapshot, it will be adjusted appropriately.
I plan to make a plan based on precise reason in any situation.