VMware Esxi Host Showing high letancy with Equallogic 6510

Helo,

Do you have SANHQ installed?  If not, you should download it from the Dell EQL support site. 

By default ESXi's network setting cache small I/Os to group them together.  This delays when data is ACK'd.  This is reported a latency by the ESXi host.

Things like Delayed ACK, Large Recieve Offload, and sometimes Large Send Offload will cause this.

Here's a a PDF which describes these and other common causes for ESXi reported latency.

http://en.community.dell.com/techcenter/extras/m/white_papers/20434601.aspx

I have found that in over 90% of the cases, these settings resolve those errors and promote better overall performance.

Regards

do we have updated copy of this document about Esxi 5.5.

second,, first script  is not working ..do we need to run some selected part of script

esxcli storage nmp satp set --default-psp=VMW_PSP_RR --satp=VMW_SATP_EQL ; for i in `esxcli storage nmp device list | grep EQLOGIC|awk '{print $7}'|sed 's/(//g'|sed 's/)//g'` ; do esxcli storage nmp device set -d $i --psp=VMW_PSP_RR ; esxcli storage nmp psp roundrobin deviceconfig set -d $i -I 3 -t iops ; done

re: 5.5   There are no changes for ESXi v5.5.   The document still applies

Re: Script.  What error message do you get?

Try this script.  Only downside is that you will have to reboot afterwards for it to become effective.

esxcli storage nmp satp rule add -s "VMW_SATP_EQL" -V "EQLOGIC" -M "100E-00" -P "VMW_PSP_RR" -O "iops=3" 

What VMware license do you have?   If you have Enterprise or Enterprise+ you can install the Dell MEM kit.  That automatically optimizes MPIO for EQL devices.

Hello,

Did you go through the other settings in the PDF I included?   Things like Delayed ACK, Large Receive Offload disabled,  making sure that your VMs aren't sharing a single Virtual SCSI adapter when you have multiple VMDKs (or Raw Device Mapped LUNs too)    

These are common causes of ESXi reporting latency.

Sometimes when you disable DelayedACK, it shows as off, but if you haven't rebooted, it likely is still in effect on the established ISCSI connections.

You can also open a support case, they'll gather diags, SANHQ archive, and VM support files to look for common problems like the Delayed ACK issue.   They will also check the network stats to make sure that the switches aren't causing issues either.

Regards,

First, when you refer to latency issues, are you referring to alerts from ESXi?

2nd, what items have you left out in the best practices guide?

3rd, I would make sure the DCB is disabled on the MXL switch.

4th, Do you have SANHQ installed?   If so, what latency is being reported there?

5th, do you have a support case open on this?

Hi team,

still facing the Letancy issue after doing almost all the changes as per the Best practice doc and mentioned in this thread.

final check on connectvity

PS6510 with two 10 G ports on each controller. Port 1 from controller A connected to MXLswicth B1 and other port of same controller connected to MXl b2. same for the other controller also.

MXL are stacked with stacking cable.