VxRail: Admin gets a pop-up Window asking to retry when using a ISO stored on the vSAN datastore[1]

           

   Article Number:     482658                                   Article Version: 3     Article Type:    Break Fix 
   

 


Product:

 

VSPEX BLUE Appliance Software

 

Issue:

 

 

When accessing a .ISO stored on the vSAN datastore from a VM in the VSPEX BLUE cluster, admin may get the following pop-up message:   
   
    User-added image   
   
    Also, vmkernel.log show entries like the following:   
   
    2016-04-28T14:27:28.917Z cpu3:260980)WARNING: Heap: 4154: Heap_Align(dom-Client-heap, 1048584/1048584 bytes, 8 align) failed.  caller: 0x41803857f4dc     
      2016-04-28T14:27:35.874Z cpu16:37279)World: 14302: VC opID 59e746ce-f9 maps to vmkernel opID f9b1dfc6     
      2016-04-28T14:27:37.108Z cpu12:1821456)Vol3: 2790: Failed to get object 28 type 2 uuid 560ae195-04cc4767-7ba0-001e67dc4b28 FD 4 gen 1 :Out of memory     
      2016-04-28T14:27:37.108Z cpu12:1821456)Vol3: 2790: Failed to get object 28 type 2 uuid 560ae195-04cc4767-7ba0-001e67dc4b28 FD 4 gen 1 :Out of memory     
      2016-04-28T14:27:37.108Z cpu12:1821456)Vol3: 2790: Failed to get object 28 type 2 uuid 560ae195-04cc4767-7ba0-001e67dc4b28 FD 4 gen 1 :Out of memory     
      2016-04-28T14:27:37.108Z cpu12:1821456)Vol3: 2790: Failed to get object 28 type 2 uuid 560ae195-04cc4767-7ba0-001e67dc4b28 FD 4 gen 1 :Out of memory     
      2016-04-28T14:27:37.109Z cpu21:1821456)Vol3: 2790: Failed to get object 28 type 2 uuid 560ae195-04cc4767-7ba0-001e67dc4b28 FD 4 gen 1 :Out of memory     
      2016-04-28T14:27:37.109Z cpu21:1821456)Vol3: 2790: Failed to get object 28 type 2 uuid 560ae195-04cc4767-7ba0-001e67dc4b28 FD 4 gen 1 :Out of memory     
      2016-04-28T14:27:37.109Z cpu21:1821456)Vol3: 2790: Failed to get object 28 type 2 uuid 560ae195-04cc4767-7ba0-001e67dc4b28 FD 4 gen 1 :Out of memory
                                                           

 

 

Cause:

 

 

This happens if the dom-Client-heap is exhausted.   
   
     
                                                           

 

 

Resolution:

 

 

To work around these errors, the dom-Client-heap needs to be expanded on all hosts in the cluster. After that, the access to the ISO files from any VM should work without any issues.   
   
    Admin needs to apply the change to dom-Client-heap on each host and reboot using the steps below:   
   
    To change the default size of the DomClientHeap:   
    ~ # esxcfg-advcfg -s 512 /VSAN/DomClientHeapSize   
   
    To save this setting across reboots:   
    ~ # /sbin/auto-backup.sh   
   
    To verify the change has persisted across a reboot, run this command:   
    ~ # esxcfg-advcfg -g /VSAN/DomClientHeapSize