Azure Stack Storage Servics role server high memory usage (ACS).

%3CLINGO-SUB%20id%3D%22lingo-sub-1210458%22%20slang%3D%22en-US%22%3EAzure%20Stack%20Storage%20Servics%20role%20server%20high%20memory%20usage%20(ACS).%3C%2FLINGO-SUB%3E%3CLINGO-BODY%20id%3D%22lingo-body-1210458%22%20slang%3D%22en-US%22%3E%3CP%3E%3CSTRONG%3ERoot%20Case%2C%3CBR%20%2F%3E%3C%2FSTRONG%3ESince%201908%2C%20Windows%20Server%202019%20introduced%20new%20sync-redirect%20feature%20for%20SMB%20to%20ensure%20client%20connecting%20to%20CSV%20owner%20node%20to%20reduce%20network%20hop%20on%20server%20side%2C%20this%20requires%20SMB%20server%20to%20return%20the%20owner%20node%20IP%20addresses%20on%20client%20TreeConnect%20call.%20In%20Azure%20Stack%20configuration%2C%20this%20will%20return%20all%203%20IP%20address%20of%20that%20owner%20node%20(1%20infrastructure%20network%2C%202%20storage%20network).%20Because%20ACS%20VMs%20are%20multi-homed%20and%20can%20connect%20to%20host%20via%20storage%20network%2C%20this%20triggers%20a%20following%20SMB%20witness%20registration%20RPC%20call%20and%20all%20its%20successive%20retries%20fail%20and%20leak%20a%20calling%20context%20every%20time.%3C%2FP%3E%3CP%3E%26nbsp%3B%3C%2FP%3E%3CP%3E%3CSTRONG%3ESolution%2C%3C%2FSTRONG%3E%3C%2FP%3E%3CP%3EWith%20the%202020%20release%20this%20issue%20will%20be%20permanently%20fixed.%20Until%20you%20can%20restart%20the%20ACS%20server%20by%20using%20Infrastructure%20Roles%20%26gt%3B%20Storage%20Service.%20Restart%20servers%20one%20by%20one.%20Wait%20at%20least%2015%20minutes%20to%20restart%20second%20server.%26nbsp%3B%3CBR%20%2F%3E%3C%2FP%3E%3C%2FLINGO-BODY%3E
Occasional Contributor

Root Case,
Since 1908, Windows Server 2019 introduced new sync-redirect feature for SMB to ensure client connecting to CSV owner node to reduce network hop on server side, this requires SMB server to return the owner node IP addresses on client TreeConnect call. In Azure Stack configuration, this will return all 3 IP address of that owner node (1 infrastructure network, 2 storage network). Because ACS VMs are multi-homed and can connect to host via storage network, this triggers a following SMB witness registration RPC call and all its successive retries fail and leak a calling context every time.

 

Solution,

With the 2020 release this issue will be permanently fixed. Until you can restart the ACS server by using Infrastructure Roles > Storage Service. Restart servers one by one. Wait at least 15 minutes to restart second server. 

0 Replies