I'm setting up a new vSphere 6.5 cluster and starting to provision my first machines. I've set up a template for future guests, which uses EFI instead of BIOS. Upon rebooting the guests, they fail to start properly.
The issue is identical to this one I found on the KB: https://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=21461...
I have followed the steps and set my configuration accordingly. Here is the output from esxcli storage core path list:
iqn.1998-01.com.vmware:esx-mk1-58d288d3-00023d000003,iqn.2015-11.com.hpe:storevirtual.rpl,t,3-naa.6000eb3b01323a960000000000001bf0
UID: iqn.1998-01.com.vmware:esx-mk1-58d288d3-00023d000003,iqn.2015-11.com.hpe:storevirtual.rpl,t,3-naa.6000eb3b01323a960000000000001bf0
Runtime Name: vmhba64:C0:T0:L10
Device: naa.6000eb3b01323a960000000000001bf0
Device Display Name: LEFTHAND iSCSI Disk (naa.6000eb3b01323a960000000000001bf0)
Adapter: vmhba64
Channel: 0
Target: 0
LUN: 10
Plugin: NMP
State: active
Transport: iscsi
Adapter Identifier: iqn.1998-01.com.vmware:esx-mk1-58d288d3
Target Identifier: 00023d000003,iqn.2015-11.com.hpe:storevirtual.rpl,t,3
Adapter Transport Details: iqn.1998-01.com.vmware:esx-mk1-58d288d3
Target Transport Details: IQN=iqn.2015-11.com.hpe:storevirtual.rpl Alias= Session=00023d000003 PortalTag=3
Maximum IO Size: 131072
iqn.1998-01.com.vmware:esx-mk1-58d288d3-00023d000004,iqn.2015-11.com.hpe:storevirtual.rpl,t,3-naa.6000eb3b01323a960000000000001bf0
UID: iqn.1998-01.com.vmware:esx-mk1-58d288d3-00023d000004,iqn.2015-11.com.hpe:storevirtual.rpl,t,3-naa.6000eb3b01323a960000000000001bf0
Runtime Name: vmhba64:C1:T0:L10
Device: naa.6000eb3b01323a960000000000001bf0
Device Display Name: LEFTHAND iSCSI Disk (naa.6000eb3b01323a960000000000001bf0)
Adapter: vmhba64
Channel: 1
Target: 0
LUN: 10
Plugin: NMP
State: active
Transport: iscsi
Adapter Identifier: iqn.1998-01.com.vmware:esx-mk1-58d288d3
Target Identifier: 00023d000004,iqn.2015-11.com.hpe:storevirtual.rpl,t,3
Adapter Transport Details: iqn.1998-01.com.vmware:esx-mk1-58d288d3
Target Transport Details: IQN=iqn.2015-11.com.hpe:storevirtual.rpl Alias= Session=00023d000004 PortalTag=3
Maximum IO Size: 131072
As per the article, I have taken the maximum IO size of 131072 (bytes) and set the 'Disk.DiskMaxIOSize' setting on each of my hosts to 128 (kilobytes). Output of esxcli system settings advanced list -o "/Disk/DiskMaxIOSize" | grep Int\ Value | grep -v Default'
Int Value: 128
Here's a snippet of vmware.log for a guest that fails to start:
2017-02-14T15:15:16.849Z| vcpu-0| I125: Guest: About to do EFI boot: Windows Boot Manager
2017-02-14T15:15:16.886Z| vcpu-0| I125: HBACommon: First write on scsi0:0.fileName='/vmfs/volumes/58986c57-beebfb62-20dc-941882e61b50/dc-mk1/dc-mk1.vmdk'
2017-02-14T15:15:16.886Z| vcpu-0| I125: DDB: "longContentID" = "9183ad927c82a0f735b6ad050dce9686" (was "b8e64cc6686f60edd799a37495b96386")
2017-02-14T15:15:16.939Z| vcpu-0| I125: DISKLIB-CHAIN : DiskChainUpdateContentID: old=0x95b96386, new=0xdce9686 (9183ad927c82a0f735b6ad050dce9686)
2017-02-14T15:15:20.002Z| vmx| I130: Vigor_ClientRequestCb: dispatching Vigor command 'Ethernet.queryFields'
2017-02-14T15:15:38.459Z| vcpu-0| I125: Tools: Tools heartbeat timeout.
There is a consistent 22 seconds between "about to do EFI boot" and "tools heartbeat timeout" when the guest fails to start.
I'm keen to resolve this ASAP so I can commission my new hardware in earnest - any ideas?
Bump
Sorry for the delay in responding... I was on vacation.
Can you please reproduce the problem again and then collect the ESXi host's vmkernel.log and the corresponding vmware.log for that run, and attach the pair of files to a reply in this thread? (Just use the Attach link in the lower-right corner when composing your reply.) That should provide more information about whether there's still a problem with the maximum IO size and possibly other places to look to understand what's happening.
Thanks,
--
Darius
Did you ever resolve this?
I got a StoreVirtual 3200 which has the same symptoms.. EFI takes from 1m til 12 hours to boot. In BIOS mode it works fine.
esxcli storage core path list gives Maximum IO Size: 131072 on the iscsi software adapter, though it should support more.
Im not sure if its the storage or vSphere..