Have you tried running all four tests?
RealLife and Random will give you a better idea of how your system will perform in "Real Life"
What block size are you using for your tests? It doesn't look like the 32K block used in OpenPerformance.
1505 IOPS / 747 MBReads/S
If you are in fact getting 747MB/s thats not bad at all.
Try searching for similar results in the Open unofficial storage performance thread
thats too bad!!!
I will open a case next week and will keep you posted if i reach anything
Many thanks for the update
What issue are you seeing? Can you post your OpenPerformanceTest results? Did these issues develop after a firmware upgrade?
has anybody a solution or update. Same Problem with ESXI 5 U1 and Cisco.
I encountered a different problem when upgrading from ESX 4.1 to ESXi 5 U1.
RealLife,Max Throughput-50% and Random tests all showed expected results. Max Throughput-100% on the other hand had horrible results. After a lot of troubleshooting I indentified the Intel quad nic as the culprit.
The following fix resolved the issue:
esxcfg-module -s "InterruptThrottleRate=0,0,0,0" igb
Thanks for the update. And what was the resolution for problem in you`re first threat? The kb is for intel nics.
You have a PM.
Can please anybody with the same problem then _VR_ with a new Equallogic look at the Switch if FlowControl is up. At the moment it looks like flowcontrol never comes up on the new PS4100.
This is an output when it`s not running, FlowControl oepr must be on:
Switch# show flowcontrol interface gigabitEthernet 5/5
Port Send FlowControl Receive FlowControl RxPause TxPause
admin oper admin oper
--------- -------- -------- -------- -------- ------- -------
Gi5/5 off off desired off 0 0
Is there any update on this situation? I am about to place an order for either a 4100X or a 4100XV and wonder if I am going to experience the same problems. My three hosts are IBM x3650 M2 boxes with two built-in Broadcom NIC ports and a quad-port Intel 82571EB NIC.
I will probably replace my current physically-dedicated SAN switches with a pair of HP 2510-24G units.
No, still not fiexed at my site. But cause is unknown.
Real live Performance is OK. But read performance is ----
Nachricht geändert durch alex555550
I'm having the exact same issue you're seeing. Alex - I've shot you a PM with more details...
So I've been working on this a lot today and I've made a little progress. Equallogic support sent me this doc:
Specifically it mentions the heartbeat VMK for iSCSI. I did not have this configured (I am using the MEMs). For a while my retransmits in SAN HQ have been above normal (<1% but above 0.1% where it should be <0.1%). I was attributing this to the flowcontrol issue that Alex pointed out. After I added the Storage Heartbeat VMK, my retransmits fell to <0.1% and my transfer rate is much higher. I would suggest trying this if you haven't already.
On a side note, I have a replica environment that has a PS6100 and Dell R710 hosts with Intel PT quad-port cards in it for failover (I'm using the same in the primary except it's a PS6500). That site has the same switches (Cisco 3750) but a slightly older firmware (12.2.55 vs 12.2.58). This site is reporting proper flowcontrol on for the storage ports. I'm not sure if the switch firmware has anything to do with it but I plan on opening up a TAC case soon to see what they say.
thanks for the update. I`ll have this Problem with all my Cisco`s.
ok cisco finaly found an bug in IOS.
CISCO BUG ID: CSCty55093