Not possible to extend HAImage over ~32 Tb

[arifurinmoto]

Hello!

Our setup: 2-node cluster w/ Windows Server 2019 Datacenter, 54 Tb ReFS disk on each node configured by Storage Spaces
We have 2 HAImage devices at this disk (VSAN Free): HAImage1 30,7Tb (for fileshare) and HAImage2 4,7Tb (for shadowcopies)

We want to extend HAImage1 for additional 5 Tb but when run ExtendDevice.ps1 param($addr="127.0.0.1", $port=3261, $user="root", $password="starwind", $deviceName="HAImage1", $extendSize=5000000) we take next warning:

Code: Select all

Failed to resize device 0x000001C82B175700
-
control 0x000001C82B175700 -Reserve:"33765195776000"
-
200 Failed: operation cannot be completed.. 

So, we dont understand what prevent disk from extending

[yaroslav (staff)]

The value you are typing is in MB there.
P.S. Be careful with ReFS https://community.spiceworks.com/t/plea ... san/774118

[arifurinmoto]

The value you are typing is in MB there.

Yes, $extendSize=5000000 ~5 Tb while we have ~20Tb of free space

[yaroslav (staff)]

Please tell me more about the system and share the script. What hypervisor do you use? Is it a Windows-based application or CVM (is it a white UI or a dark one)?

P.s. Please also note that we do not generally recommend devices bigger than 6 TB (https://www.starwindsoftware.com/best-p ... practices/). The bigger the device is the longer it takes to synchronize it. Full synchronization for large devices may overlap with productive and backup hours, causing performance impact or synchronization might drop due to other processes causing latency to go too high. As a result, such a synchronization may never finish.

[arifurinmoto]

We do not use hypervisor. We do not have CSV. We use volumes on HAImage disks as Cluster disk on our File Server Cluster.

CreateHA_2 :

Code: Select all

param($addr="10.35.0.1", $port=3261, $user="root", $password="starwind",
	$addr2="10.35.0.2", $port2=$port, $user2=$user, $password2=$password,
#common
	$initMethod="Clear",
	$size=1000,
	$sectorSize=4096,
	$failover=0,
	$bmpType=1,
	$bmpStrategy=0,
#primary node
	$imagePath="My computer\E\SW",
	$imageName="fs1share",
	$createImage=$true,
	$storageName="",
	$targetAlias="targetfs1",
	$autoSynch=$true,
	$poolName="pool1",
	$syncSessionCount=1,
	$aluaOptimized=$true,
	$cacheMode="wt",
	$cacheSize=1000,
	$syncInterface="#p2=10.33.0.2:3260" -f $addr2,
	$hbInterface="#p2=10.34.0.2:3260,10.35.0.2:3260",
	$createTarget=$true,
	$bmpFolderPath="",
#secondary node
	$imagePath2="My computer\E\SW",
	$imageName2="fs2share",
	$createImage2=$true,
	$storageName2="",
	$targetAlias2="targetfs2",
	$autoSynch2=$true,
	$poolName2="pool1",
	$syncSessionCount2=1,
	$aluaOptimized2=$true,
	$cacheMode2=$cacheMode,
	$cacheSize2=$cacheSize,
	$syncInterface2="#p1=10.33.0.1:3260" -f $addr,
	$hbInterface2="#p1=10.34.0.1:3260,10.35.0.1:3260",
	$createTarget2=$true,
	$bmpFolderPath2=""
	)

ExtendDevice:

Code: Select all

param($addr="127.0.0.1", $port=3261, $user="root", $password="starwind", $deviceName="HAImage1", $extendSize=5000000)

[yaroslav (staff)]

$syncInterface="#p2=10.33.0.2:3260" -f $addr2,

This line is worrying see more about script syntax viewtopic.php?f=5&t=6852&p=37208&hilit=HINT8#p37208.
Did it create the device?
You can create a larger device with NotSynchronize option (it skips synchronization).

[arifurinmoto]

About syntax
instead of
$syncInterface="#p2=10.33.0.2:3260" -f $addr2,

it could be
$syncInterface="#p2=10.33.0.2:3260,$addr2:3260" ,

or
$syncInterface="#p2={0}:3260" -f 10.33.0.2,$addr2,
?

[yaroslav (staff)]

Do not use $addr.

[arifurinmoto]

So,
$syncInterface="#p2=10.33.0.2:3260,10.35.0.2:3260" ,

equals
$syncInterface="#p2={0}:3260" -f 10.33.0.2,10.35.0.2,
?

[yaroslav (staff)]

No, it's slightly different. See viewtopic.php?f=5&t=6852&p=37208&hilit=HINT8#p37208.

[arifurinmoto]

Sorry, I didn't find exactly what the difference between these two options is.

[yaroslav (staff)]

The first one works.

[arifurinmoto]

Thanks for the last tip! We've fixed our scripts for the future.

But our main problem is still unresolved.
We decided to split our storage into two parts and create a new HAImage.
And now we are facing the same problem:

Code: Select all

PS C:\Windows\system32> C:\SW_scripts\CreateHA_2_yet_another_disk.ps1
Request to  FS1.DOMEN ( 10.35.0.1 ) : 3261
-
control ImageFile -CreateImage:"My computer\E\SW\fs1share.img" -Size:"1000000000" -Flat:"True" -DeferredInit:"True" -Password:"starwind"
-
200 Failed: There is not enough space on the disk.. 

But we have >18Tb free space on disk E.
What could prevent the creation of a new HAImage or extension of an existing one?

[yaroslav (staff)]

Do both replication partners have enough space?
Could you please try creating a small device (1 GB) and expanding it afterward in several hops (e.g., add 3 TB, 6TB, and finally 9TB)?

[arifurinmoto]

Both partners have > 18Tb of free space.

We create small image 1Gb - OK
We extend it for 1 Gb - OK
We extend it for 100 Gb - OK
We extend it for 1 Tb - error There is not enough space on the disk..