How do you identify the source of reads for a workload in ONTAP 9?
Applies to
- ONTAP 9
- Cloud Volume ONTAP 9
Answer
- The
workload_volume
CM object has a counter calledread_io_type
which can be used to identify the sources of reads. - Use the commands to get the output:
Cluster::> set -privilege diagnostic
Cluster::*> statistics start -object workload_volume -counter read_io_type
Cluster::*> statistics show
Cluster::*> statistics stop
- The following are all the different read sources:
Counter | Description |
---|---|
read_io_type.cache |
Reads from system RAM |
read_io_type.pmem |
Read from persistent memory |
read_io_type.ext_cache |
Reads from Flash Cache |
read_io_type.disk |
Reads from an HDD aggregate |
read_io_type.bamboo_ssd |
Reads from an SSD aggregate |
read_io_type.hya_hdd |
Reads from Flash Pool HDD RAID group |
read_io_type.hya_cache |
Reads from Flash Pool SSD RAID group |
read_io_type.hya_non_cache |
Not used |
read_io_type.cloud |
Reads from composite aggregate (FabricPool cold tier) |
read_io_type.fc_miss |
Reads from FlexCache origin and remote retrieves like a remote FG member (missed cache lookup) |
read_io_type.cloud_s2c |
Non FabricPool cloud reads such as restores from SnapMirror to Cloud solution |
Additional Information
How to check if FlashPool or FlashCache is effectively caching reads