Yet another ConfigMgr consulting day, yet another interesting find 🙂
We, meaning me and Jordan Benzing, were doing a ConfigMgr Health Check for a decent size ConfigMgr customer (65K clients, of which about 8K client were connecting via Cloud Management Gateway, CMG). About halfway into the health check we discovered that 45 of about 350 packages targeted for the Cloud DP were not distributing correctly.
The Monitoring node in ConfigMgr showed the following error: Cloud Content Manager encountered an empty download address from content upload for package
But, the PkgXferMgr.log file gave the real clue to the problem: Failed to generate contentinfo for the package. BranchCache feature might not be turned on.
It turned out BranchCache was turned on in the environment, but that the BranchCache publication cache had hit its limit. The default size setting is 1 percent of the hard drive (C: by default).
Once learning about the issue, the fix was not terribly complicated. On the DP, flush the BranchCache publication cache, move it to a different disk, and set the size to 5 GB. For good measure we also restarted the BranchCache service.
Here are the commands we run in an elevated PowerShell prompt:
$NewHashFolder = "E:\BCPublicationCache" $NewHashSize = 5GB New-Item -Path $NewHashFolder -ItemType Directory $BCCache = Get-BCStatus Set-BCCache -Path $BCCache.HashCache.CacheFileDirectoryPath -MoveTo $NewHashFolder -Force $BCHashCache = Get-BCHashCache $BCHashCache | Set-BCCache -SizeBytes $NewHashSize -Force
If you're on older DPs (Windows Server 2008 R2), run this instead:
netsh branchcache flush md E:\BCPublicationCache netsh branchcache set publicationcache directory=E:\BCPublicationCache netsh branchcache set publicationcachesize 5368709120
Happy Deployment / Johan