Resolution to previous ImageBuilder process issue for Service Fabric Linux Clusters


A previous issue (Known Issue for Service Fabric Linux Clusters) which had impacted the ImageBuilder process on Azure based Service Fabric Linux clusters has been resolved. The resolution to this issue comes from fixes introduced in the Service Fabric 6.4.644 runtime and an update to the Ubuntu xenial rssh package (2.3.4-4+deb8u2ubuntu0.16.04.2).

Impacted Customers 

Any Service Fabric Linux cluster which required a custom script extension added to the Azure Resource Manager (ARM) template in order to mitigate the ImageBuilder process issue.


We had previously recommended that users impacted by this issue, add a custom script extension to their Azure Resource Manager (ARM) template. This custom script extension can now be removed from the ARM template if the cluster is running version 6.4.644+ of the Service Fabric runtime. Once the custom script extension is removed and the template has been redeployed, no additional action will be required.

Remove the below custom script extension from the extensions section of your Azure Resource Manager (ARM) template.


"name": " [concat(variables('vmNodeType0Name'),'_rssh_2_4_4-4-fix')]",

     "properties": {

         "publisher": "Microsoft.Azure.Extensions",

         "settings": {

             "fileUris": [




         "typeHandlerVersion": "2.0",

             "autoUpgradeMinorVersion": true,

             "protectedSettings": {

                 "commandToExecute": "./"


             "type": "CustomScript"



Please reach out to us if you need any assistance through the Azure Portal Help. In case it is needed, here are your general support options for Service Fabric:

Comments (1)
  1. Cloudbrick says:

    I’m tring to deploy a SF Cluster on Ubuntu 16.04 from 3 days now and with no luck. All that i have after 3 hrs of waiting its a huge amount of dump files in the storage and no nodes available. The curent status its Deploying.

Comments are closed.

Skip to main content