what am i doing wrong? VSan Free 2 node hpyer-v/ms cluster

[cwilson@csjssm.ca]

Eventually plan on getting the paid version of Vsan. But as a charity $ is tight so I'm setting up with the free version for now

I have 2 nodes. HV1 and HV2
both have 3 teamed networks (primary(2x1gb)/heartbeat(2x1gb)/replication(2x10gb))
MS Failover clustering is installed
HyperV is installed
I have created on HV1 3tb HA with a partner on HV2
I have created a second on HV2 3tb HA with a partner on HV1
I have also created a 2gb HA on HV1 with a partner on HV2 to use a Quorum witness

MPIO is installed
Now I'm at the IScsi stage (this is different to me as I come from a FC background, ie zoning, etc)

I have added both hosts into the discovery in the ISCSI Initiator and it comes up with 7 targets
do i pick the iscsi targes on the opposite host for these or do i have to connect all the targets on both systems?

is it best practice to have 1 6tb volume and do 2way sync or 2 3tb with a mirror on the other server and have

[yaroslav (staff)]

both have 3 teamed networks (primary(2x1gb)/heartbeat(2x1gb)/replication(2x10gb))

Please destroy the teaming. We recommend using the dedicated links. See more at https://www.starwindsoftware.com/best-p ... practices/
Also, I'd recommend using iSCSI and sync of the same bandwidth. How many network cards do you have there? A screenshot of get-netadapter outputs will be much appreciated.

do i pick the iscsi targes on the opposite host for these or do i have to connect all the targets on both systems?

Yes, you connect both hosts, one over 127.0.0.1 another over iSCSI connections as described here https://www.starwindsoftware.com/resour ... 0e77312268
Quorum is connected over 127.0.0.1 if that's hyperconverged scenario (VMs and storage are on the same servers).

is it best practice to have 1 6tb volume and do 2way sync or 2 3tb with a mirror on the other server and have

Volumes are recommended to be less than 6 TB. It is better to create 2 3 TB volumes to have some load balancing between the targets. Make sure to have devices of higher priority on one node (i.e., create devices on one side and replicate them to another).
More reading on priorities https://forums.starwindsoftware.com/vie ... 31&p=32173 good script https://forums.starwindsoftware.com/vie ... p+3#p31505

[cwilson@csjssm.ca]

OK i will work on removing teaming from some of the interfaces.(will it work without breaking the teams??)

currently, there are 2 nics in the system a Quad 1gbe and a dual 10gbe

right now the way it is configured

servers.jpg (111.81 KiB) Viewed 4090 times

with the juniper switches eliminating the single point of failure as the teamed nics are spanned across the 2 switches

[yaroslav (staff)]

Hi,

It will be supported without a team. Why not connecting the servers directly over 10 GBE?

[cwilson@csjssm.ca]

when i try to add the local iscsi targets on the 2nd host I get an error that the service is unavailable(there was no issue adding it to the 1st host

Screenshot_1.jpg (191.75 KiB) Viewed 4070 times

Screenshot_2.jpg (103.5 KiB) Viewed 4070 times

this is what I have in the discovery

Screenshot_3.jpg (81.44 KiB) Viewed 4070 times

this is the directions I am following
https://www.starwindsoftware.com/resour ... rver-2016/

[yaroslav (staff)]

Are HA devices synchronized?
Were you able to add both partner connections on the 1st server?
I noticed that you have 2 iSCSI interfaces in the same subnet. Please use different subnets. Also, do not reuse any of them for synchronization or VM connections.
You need to set <iScsiDiscoveryListInterfaces value=”0”/> to 1. Stop the service on one side -> go to StarWind.cfg -> copy the .cfg -> Make the change -> save it -> start the service -> wait for sync to be over -> repeat for the 2nd server.

[cwilson@csjssm.ca]

the 2 10gb ports from each server have been directly connected (eliminating the switches) right now I'm in the process of removing all the images/targets that were created when the nics were teamed as none of the addressing was still valid

[yaroslav (staff)]

You do not need to remove the images. You can add the new IP addresses for the existing ones.
Please let me know if assistance is required.

[cwilson@csjssm.ca]

I have already removed the images(they were all misnamed/etc from testing things out) and were all blank
but now when I have started to make new ones its staying at 0% sync while starting to make the 1st HA set
live address node 1 192.168.1.231 and node 2 192.168.1.232 (they can ping each other)
heart beat add node 1 10.1.1.231 and node 2 10.1.1.232 (they can ping each other)
iscsi interface node 1 int 1 10.10.10.11 node 2 int 1 10.10.10.21 (they can ping each other)
iscsi interface node 1 int 2 10.10.11.11 node 2 int 2 10.10.11.21 (they can ping each other)




param($addr="192.168.1.231", $port=3261, $user="root", $password="starwind",
$addr2="192.168.1.232", $port2=$port, $user2=$user, $password2=$password,
#common
$initMethod="Clear",
$size=30000,
$sectorSize=512,
$failover=0,
#primary node
$imagePath="h:\starwind",
$imageName="witness_HV11",
$createImage=$true,
$storageName="",
$targetAlias="witness_HV1",
$autoSynch=$true,
$poolName="pool1",
$syncSessionCount=1,
$aluaOptimized=$true,
$cacheMode="wb",
$cacheSize=128,
$syncInterface="#p2=10.10.10.11:3260" -f $addr2,
$hbInterface="#p2=10.1.1.232:3260",
$createTarget=$true,
#secondary node
$imagePath2="h:\starwind",
$imageName2="witness_HV12",
$createImage2=$true,
$storageName2="",
$targetAlias2="witness_HV2",
$autoSynch2=$true,
$poolName2="pool1",
$syncSessionCount2=1,
$aluaOptimized2=$false,
$cacheMode2=$cacheMode,
$cacheSize2=$cacheSize,
$syncInterface2="#p1=10.10.10.21:3260" -f $addr,
$hbInterface2="#p1=10.1.1.231:3260",
$createTarget2=$true
)

[yaroslav (staff)]

Check this script https://forums.starwindsoftware.com/vie ... p+3#p31505
User here https://forums.starwindsoftware.com/vie ... p+3#p32679 had exact the same problem.

[cwilson@csjssm.ca]

I think i have done all the steps required but still, it is showing and unable to connect to partner

[cwilson@csjssm.ca]

files.zip

(16.99 KiB) Downloaded 153 times

here is the ha2 create i am using and config file from both hosts

[yaroslav (staff)]

The primary node in the script is refering to the IP addresses of the node1. In other words, it is usin the IPs of itself.
Please type there the IP addresses of the partner.