1. You should Aim to keep your FA utilization below,70% but aim for 50%. This is more for redundancy than limits. You need to ensure you have enough spare utilization in case another path fails and you need to operate in a degraded scenario keeping good response times.
2. It depends. 15K FC drives can do about 150 IOPS each at a reasonable response time, but depending on the raid config and access pattern RW/Ratio your mileage will vary. With Raid 1 a write will do 2 IOps so you have a one to each mirror of the drive. See the following thread for calculating IOPS capability
1. symcli command can't list the FA utilization which only SPA can do. symstat can provide the IO/sec and KBytes/Sec for general status about the FA ports. But you can see compare each port and make it balance.
For calculate maximum IO for you thin meta device. You should at least know your IO profile( Read/Write ratio), like Paul said, the Raid write penalty should be considered. As your multiple hyper shares same data device on two physical disk. The answer is it depends.
Device Performance Statistics. symstat can display performance statistics such as I/O requests via front-end adapters and throughput on devices. For example, for sample intervals of 120 seconds and a sample count of 3 on logical device DEV001 in device group prod_r1, use command:
symstat -i 250 -c 3 -g prod_r1 -ld DEV001
3. You can't specify the particular IG group through symstat command. GUI tools is required. SPA or Unisphere for VMAX
rawstorage
3 Apprentice
•
423 Posts
0
December 13th, 2013 00:00
Santosh,
in response to your questions above.
1. You should Aim to keep your FA utilization below,70% but aim for 50%. This is more for redundancy than limits. You need to ensure you have enough spare utilization in case another path fails and you need to operate in a degraded scenario keeping good response times.
2. It depends. 15K FC drives can do about 150 IOPS each at a reasonable response time, but depending on the raid config and access pattern RW/Ratio your mileage will vary. With Raid 1 a write will do 2 IOps so you have a one to each mirror of the drive. See the following thread for calculating IOPS capability
https://community.emc.com/thread/148025
3.You can use the symstat command set to monitor IOs for a group, but the preferred method is to use Unisphere for VMAX performance analyzer.
Fenglin1
4 Operator
•
2.1K Posts
0
December 15th, 2013 19:00
1. symcli command can't list the FA utilization which only SPA can do. symstat can provide the IO/sec and KBytes/Sec for general status about the FA ports. But you can see compare each port and make it balance.
2. See Monitoring Symmetrix Performance using symstat CLI for getting IOs for device level.
For calculate maximum IO for you thin meta device. You should at least know your IO profile( Read/Write ratio), like Paul said, the Raid write penalty should be considered. As your multiple hyper shares same data device on two physical disk. The answer is it depends.
Device Performance Statistics. symstat can display performance statistics such as I/O requests via front-end adapters and throughput on devices. For example, for sample intervals of 120 seconds and a sample count of 3 on logical device DEV001 in device group prod_r1, use command:
symstat -i 250 -c 3 -g prod_r1 -ld DEV001
3. You can't specify the particular IG group through symstat command. GUI tools is required. SPA or Unisphere for VMAX
ss026230
8 Posts
0
December 15th, 2013 20:00
Thanks , it was very helpful.
Thanks
Santosh S