Just did a implementation. Here is a brief summary.
1. check software iscsi installed
2. use iscsiutil -l to get the name of your initiator
3. present volumes to your initiator (better copy/paste to avoid typo!) ) on your lefthand CMC
4. use iscsiutil – a -I IP_P4000_Target -M to add target
5. run ioscan and insf, you should be able to see the new disks.
to see all the target found, use iscsiutil -pD
in case you have to change authentication method, use
iscsiutil -t authmethod None/CHAP
Multi-pathing: I couldn’t find any information on multi-pathing so I just used APA to team the NICs for iSCSI , tested by unplugging cables. It works.
Note: ESX Multi-pathing is different on version 3.x and version 4.x. on this notes, we only talk about 4.x.
If you use default VMW_SATP_DEFAULT_AA, you will end up using all the ports on both controllers for I/O. For a ALUA array, you should let OS choose the owning controller to maximize the performance.
by default, you may get this
Now, add the appropriate rule.
#esxcli nmp satp addrule –satp=”VMW_SATP_ALUA” –vendor=”HP” –model=”^P2000*” –description=”HP MSA G3 A/A with ALUA”
Reboot to activate the role. If the user cannot reboot, trigger a reclaim action:
# esxcli corestorage claiming reclaim -d <NAA_ID>
If the reclaim does not work, reboot.
Now you should see all paths as “active” and I/O will go through the owning controller.
I’m getting really motivated to start this blog because of all the things happened lately. I feel energetic and excited. Hello World!
Configuration:2 ESX servers connected to P2000 via iSCSI. 24x300GB capacity
This client had connection lost issue on their previous iSCSI array. This HP P2000 is the replacement unit.
Here is what I’m going to do in this project:
1. verify the compatibility in this environment.
2.finalize lun design with customer.
3. Configure the array and ESX servers according to best practices.
4. Benchmark the performance , then tweak accordingly.
Chapter 1. Best practices learned
1. P2000 configuration:
For applications that access both sequential and random data, use the standard optimization mode, which sets the cache block size to 32 KB. For example, use this mode for transaction-based and database update applications that write small files in random order.
For applications that access only sequential data and that require extremely low latency, use the super-sequential optimization mode, which sets the cache block size to 128 KB. For example, use this mode for video playback and multimedia post-production video- and audio-editing applications that read and write large files in sequential order
“Read ahead cache size” and “Cache optimization” should be set to “Maximum” and “super-sequential” when the application is video editing…
In order to get the maximum sequential performance from a vdisk, you should only create one volume per vdisk. Otherwise you will introduce randomness into the workload when multiple volumes on the vdisk are being exercised
2. ESX/ESXi iSCSI SAN Requirements
Configure your system to have only one VMFS datastore for each LUN. (In VMFS-3, you do not need to set accessibility.)Use RDMs for access to any raw disk.Set the SCSI controller driver in the guest operating system to a large enough queue. You can set the queue depth for the physical HBA during system setup.On virtual machines running Microsoft Windows, increase the value of the SCSI TimeoutValue parameter to allow Windows to better tolerate delayed I/O resulting from path failover.disk alignment will be automatically done when disk was created using VI client.
You must meet several requirements for your ESX/ESXi host to work properly with a SAN.
Chapter 2. Configuration step by step
reference <<Configuring VMware vSphere SoftwareiSCSI With Dell EqualLogic PS SeriesStorage>>
1. configure separate vswitch for iscsi
#esxcfg-vswitch –a vSwitch2
#esxcfg-vswitch –l will show the new vswitch2
2. add iSCSI VMKernel Ports
#esxcfg-vswitch –A iSCSI1 vSwitch2
#esxcfg-vmknic –a –i 10.10.5.173 –n 255.255.255.0 iSCSI1
#esxcfg-vswitch –A iSCSI2 vSwitch2
#esxcfg-vmknic –a –i 10.10.5.174 –n 255.255.255.0 iSCSI2
To verify the configuration enter the following command: #esxcfg-vswitch –l
To show the VMkernel ports that are assigned to the vSwitch ##esxcfg-vmknic –l
3: Assign Network Adapters
To list all of the adapters in the system run the following command:#esxcfg-nics –l
to assign vmnic1 and vmnic2 to vSwitch2
#esxcfg-vswitch –L vmnic1 vSwitch2
#esxcfg-vswitch –L vmnic2 vSwitch2
4: Associate VMkernel Ports to Physical Adapters
it HAS TO BE ONE-to-ONE relationship
using #esxcfg-vswitch –l again to list the vmnic to be removed from the association. then
#esxcfg-vswitch –p iSCSI1 –N vmnic2 vSwitch2 remove vmnic2 from iscsi1
#esxcfg-vswitch –p iSCSI2 –N vmnic1 vSwitch2 remove vmnic1 from iscsi2
5: Enable VMware iSCSI Software Initiator
6: Binding VMkernel Ports to iSCSI Software Initiator
first thing to do is to note the vmhba# of the iSCSI Software Initiator
will show “vmhba33 iscsi_vmk link-n/a iscsi.vmhba33″
we find vmknic1 and vmknic2 are iscsi
we can map each VMkernel Port to the iSCSI Software Initiator
#esxcli swiscsi nic add –n vmk1 –d vmhba33
#esxcli swiscsi nic add –n vmk2 –d vmhba33
#esxcli swiscsi nic list –d vmhba33
7: Connect to the P2000
map a volume at P2000 to all iscsi ports;
vCenter : Configuration ->storage adapters->Properties ->Dynamic Discovery -> Add target ip address, OK
8: Enabling VMware Native Multipathing – Round Robin
#esxcli nmp satp addrule –satp=”VMW_SATP_ALUA” –vendor=”HP” –model=”^P2000*” –description=”HP MSA G3 A/A with ALUA” –psp VMW_PSP_RR
#esxcli corestorage claimrule load
#esxcli corestorage claimrule run
–to be continued—