VMware ESXi host seeing low iops from E-Series array
Applies to
- E-series
- VMware ESXi
Issue
Error from VMware ESXi:
"I/O latency increased from average value of 16649 microseconds to 366839 microseconds".
Troubleshooting Steps
- Verify the number of drives used to create the Volume Group or Disk Pool.
- Calculate the IOP's based on the disk RPM.
- Example: If the disk RPM is 7200, per drive can support up to 75-100 iops.
- Formula:
number of drive in the volume group or disk pool * iops
Note: While calucuating the total iops for disk pool need to exclude the preservation drives count.
Table of the IOPS based on the disk rpm
Device Type IOPS Interface
=========== ===== ================ ===========
7,200 rpm SATA ~75-100 IOPS[2] SATA 3 Gb/s
10,000 rpm SATA ~125-150 IOPS[2] SATA 3 Gb/s
10,000 rpm SAS ~140 IOPS[2] SAS
15,000 rpm SAS ~175-210 IOPS[2] SAS
- Verify if there any errors been reported in
ISCSI-STATISTICS.CSVfile.
- Check if there are any open and queued request on drive side.
-> chall 0
chAll (Tick 1261153685) ==> 03/07/24-22:59:11
2806-A 08.74.01.00
.....Channels.....:...........Target...........:............Initiator..........:
Link :ITN :..........IOs..........:ITN :............IOs...........:.........Busy.........Idle.
Ch H/D STP Down :cnt : Open Completed Errs :cnt : Qd Open Completed Errs: Ms Ms
---- --- --- ---- :--- :----- ---------- ----- :--- :--- ----- ---------- -----: ------------ -------------
0 Drv SAS 0 : 1 : 0 42039 0 : 10 : 0 33 63446794 0: 51609897 3232649
1 Drv SAS 0 : 1 : 0 41952 0 : 10 : 21 45 54312522 0: 51465021 3377524
2 Hst iSC 0 : 1 : 0 22222954 673 : 0 : 0 0 0 0: 54842546 0
3 Hst iSC 0 : 1 : 36 22218796 462 : 0 : 0 0 0 0: 54842546 0
4 Drv iSA 0 : 0 : 0 0 0 : 1 : 0 0 315 0: 489 54842056
- Verify if there are any packet drops/errors from the network level.