svhelden

Connection from Skype to Audiocodes not working after KB4540670, fixed by disabling VMQ

Hi, we have a Skype for Business installation and multiple Audiocodes Mediant 1000 gateways. Skype server is running on Windows Server 2016 on top of Hyper-V of Windows Server 2016. Host has Intel X520 NICs. VM has used VMQ, but not SR-IOV, for years without
any issues.

Yesterday we installed KB4540670 on both the Hyper-V hosts as well as the VM. Today around noon we suddenly received reports from all our locations that PSTN calls would not work. I found that Audiocodes logs were full of messages like this:

— SRTP Error - failed decrypting RTCP packet (PacketLength=191, PayloadLength=0): wrong SSRC (11) [Code:0x3700e] [CID:53]
SRTCP_SSRC_ERROR:1  [Code:0x5004] [CID:53]

And there were corresponding events in Skype log:

An attempt to route to a gateway failed. Could not route to Gateway xxxxx, the attempt failed with response code: 503 Service Unavailable 

Uninstalling KB4540670 from the VM or moving it to another Hyper-V host did not help anything. After reboot we probably could make one or two calls but then it stopped working.

Eventually I disabled VMQ for that VM, and everything started to work in that very instant.

Could there be a compatibility issue with either VMQ or the Intel drivers in KB4540670?

Read full post . . . .

No Live Migration after KB4487026 (also not after removing it again)

Hi, we’re running a 3-node Hyper-V cluster with Windows Server 2016 and a FibreChannel SAN. This morning I installed KB4487026 (and SSU KB4485447) on the first node. 

After the first reboot, the server crashed with BSOD USER_MODE_HEALTH_MONITOR. After resetting it again, server booted fine, had new build installed (14393.2791), and seemed to work. However, it can no longer receive live migrations. All VMs failed
their migration back. For each attempt, I get this event:

Log Name:      Microsoft-Windows-Hyper-V-High-Availability-Admin
Source:        Microsoft-Windows-Hyper-V-High-Availability
Date:          2/17/2019 11:15:26 AM
Event ID:      21123
Task Category: None
Level:         Error
Keywords:      
User:          SYSTEM
Computer:      <...>
Description:
Failed to set the migration networks at the VMMS: Unspecified error (0x80004005).
Event Xml:

  
    
    21123
    0
    2
    0
    0
    0x8000000000000000
    
    10098
    
    
    Microsoft-Windows-Hyper-V-High-Availability-Admin
    <...>
    
  
  
    
      %%2147500037
      0x80004005
    
  

KB4487026 describes a known issue with SCVMM that seems related:

For hosts managed by System Center Virtual Machine Manager (SCVMM), SCVMM cannot enumerate and manage 
logical switches deployed on the host after installing the update.

I cannot perform the suggested workaround (compile two MOF files) because these MOF files are specific to SCVMM and do not exist on our servers.

Still I concluded that the issue must be related to the update and uninstalled it. That did not help though - even after removing KB4487026, Live Migration is still not possible.

I quick-migrated some VMs to the node for now. Does anyone have the same issue, or an idea?

(Live migrations from the updated node to the other nodes or between the other nodes, as well as quick migrations between all nodes, are working fine.)

Read full post . . . .

How to compact a VHDX with ReFS?

Hi, we have Exchange 2016 running on virtual Windows Server 2012 R2 on Hyper-V 2012 R2. For the databases we are using a ReFS volume, as recommended by MS. (It is a separate VHDX file that contains only one partition which is formatted with ReFS.)

The disk/volume is 4.5 TB in size, and though it never was full, the VHDX file grew to 4.5 TB. We did some cleanup and now have more than 2 TB free on the volume, but still I cannot shrink the volume to less than 4.4 TB. Also our SAN considers 4.4 TB as
allocated.

I used SDELETE and PRECOMPACT to overwrite free space with zeroes, also optimized the drive (through the GUI and through Powershell with -RETRIM option), all of that multiple times, but without success. I guess that ReFS somehow does not really write the
zeroed clusters but just marks them as unused or something like that, and the Hypervisor is not aware of that.

Is there anything I can do? (Except for creating a new VHDX and copying the files over, for which I don’t have enough space.)

Read full post . . . .

How to compact a VHDX with ReFS?

Hi, we have Exchange 2016 running on virtual Windows Server 2012 R2 on Hyper-V 2012 R2. For the databases we are using a ReFS volume, as recommended by MS. (It is a separate VHDX file that contains only one partition which is formatted with ReFS.)

The disk/volume is 4.5 TB in size, and though it never was full, the VHDX file grew to 4.5 TB. We did some cleanup and now have more than 2 TB free on the volume, but still I cannot shrink the volume to less than 4.4 TB. Also our SAN considers 4.4 TB as
allocated.

I used SDELETE and PRECOMPACT to overwrite free space with zeroes, also optimized the drive (through the GUI and through Powershell with -RETRIM option), all of that multiple times, but without success. I guess that ReFS somehow does not really write the
zeroed clusters but just marks them as unused or something like that, and the Hypervisor is not aware of that.

Is there anything I can do? (Except for creating a new VHDX and copying the files over, for which I don’t have enough space.)

Read full post . . . .

How to compact a VHDX with ReFS?

Hi, we have Exchange 2016 running on virtual Windows Server 2012 R2 on Hyper-V 2012 R2. For the databases we are using a ReFS volume, as recommended by MS. (It is a separate VHDX file that contains only one partition which is formatted with ReFS.)

The disk/volume is 4.5 TB in size, and though it never was full, the VHDX file grew to 4.5 TB. We did some cleanup and now have more than 2 TB free on the volume, but still I cannot shrink the volume to less than 4.4 TB. Also our SAN considers 4.4 TB as
allocated.

I used SDELETE and PRECOMPACT to overwrite free space with zeroes, also optimized the drive (through the GUI and through Powershell with -RETRIM option), all of that multiple times, but without success. I guess that ReFS somehow does not really write the
zeroed clusters but just marks them as unused or something like that, and the Hypervisor is not aware of that.

Is there anything I can do? (Except for creating a new VHDX and copying the files over, for which I don’t have enough space.)

Read full post . . . .

How to compact a VHDX with ReFS?

Hi, we have Exchange 2016 running on virtual Windows Server 2012 R2 on Hyper-V 2012 R2. For the databases we are using a ReFS volume, as recommended by MS. (It is a separate VHDX file that contains only one partition which is formatted with ReFS.)

The disk/volume is 4.5 TB in size, and though it never was full, the VHDX file grew to 4.5 TB. We did some cleanup and now have more than 2 TB free on the volume, but still I cannot shrink the volume to less than 4.4 TB. Also our SAN considers 4.4 TB as
allocated.

I used SDELETE and PRECOMPACT to overwrite free space with zeroes, also optimized the drive (through the GUI and through Powershell with -RETRIM option), all of that multiple times, but without success. I guess that ReFS somehow does not really write the
zeroed clusters but just marks them as unused or something like that, and the Hypervisor is not aware of that.

Is there anything I can do? (Except for creating a new VHDX and copying the files over, for which I don’t have enough space.)

Read full post . . . .

Go Que Newsroom Categories

Fatal error: Allowed memory size of 67108864 bytes exhausted (tried to allocate 36 bytes) in /home/content/36/8658336/html/goquecom/wp-includes/wp-db.php on line 1996