Protocol: SRP Operations
This statistic shows SRP operations/second (SRP IOPS) requested by initiators to Oracle ZFS Storage Appliance. Various useful breakdowns are available: to show the initiator, target, type, and latency of the SRP I/O.
Example
See Protocol: iSCSI Bytes for an example of a similar statistic with similar breakdowns.
When to Check SRP Operations
SRP operations/sec can be used as an indication of SRP load.
Use the latency breakdown when investigating SRP performance issues, especially to quantify the magnitude of the issue. This measures the I/O latency component for which Oracle ZFS Storage Appliance is responsible, and displays it as a heat map so that the overall latency pattern can be seen, along with outliers. If the SRP latency is high, drill down further on latency to identify the client initiator, the type of operation and LUN for the high latency, and check other statistics for both CPU and Disk load to investigate why the appliance is slow to respond; if latency is low, the appliance is performing quickly, and any performance issues experienced on the client initiator are more likely to be caused by other factors in the environment, such as the network infrastructure, and CPU load on the client itself.
The best way to improve performance is to eliminate unnecessary work, which can be identified through the client initiator, lun, and command breakdowns.
SRP Operations Breakdowns
Table 5-51 Breakdowns of SRP Operations
| Breakdown | Description |
|---|---|
|
initiator |
SRP client initiator. |
|
target |
Local SCSI target. |
|
project |
The project for this SRP request. |
|
lun |
The LUN for this SRP request. |
|
type of operation |
SRP operation type. This shows how the SCSI command is transported by the SRP protocol, which can give an idea to the nature of the I/O. |
|
command |
SCSI command sent by the SRP protocol. This can show the real nature of the requested I/O (read/write/sync-cache/...). |
|
latency |
A heat map showing the latency of SRP I/O, as measured from when the SRP request arrived on Oracle ZFS Storage Appliance from the network, to when the response is sent; this latency includes the time to process the SRP request, and to perform any disk I/O. |
|
offset |
A heat map showing the file offset of SRP I/O. This can be used to identify random or sequential SRP IOPS. Use the Disk I/O operations statistic to check whether random SRP IOPS maps to random Disk IOPS after the LUN and RAID configuration has been applied. |
|
size |
A heat map showing the distribution of SRP I/O sizes. |
These breakdowns can be combined to produce powerful statistics. For example: Protocol: SRP operations per second of command read broken down by latency (to examine latency for SCSI reads only).
Further Analysis
-
Protocol: SRP Bytes for SRP bytes/second
-
Protocol: SRP Average Latency for average latency per second
-
Cache: ARC Accesses to learn how well an SRP read workload is returning from cache
-
Disk: I/O Operations for the back-end disk I/O caused