...
BugZero found this defect 1071 days ago.
This article documents advanced settings to optimize Veeam Cloud Connect to handle many connections. Numerous optimizations to the Cloud Connect framework on both the service provider and the tenant sides have been introduced to increase its scalability (including the number of concurrent tenant jobs per Cloud Connect server) and reliability at scale. The Veeam Backup & Replication 12 release was tested with 1000 concurrent workloads per Cloud Connect instance. Reaching this number may require applying additional tweaks provided in this article. One recommended approach for building scalable architecture is using the pod design.
The registry values below are to be set on each server in the infrastructure. LogDirectory Defines the default log storage directory. (For more information, see KB1825) Key Location: HKLM\SOFTWARE\Veeam\Veeam Backup and Replication\Value Name: LogDirectoryValue Type: String Value (REG_SZ)Value Data: <path> Recommendation: Specify a folder location with sufficient free space to ensure Note: The path provided must be to a folder on a drive of the server without the trailing slash. The path cannot be a remote location or mapped drive. Restart not required, but recommended to ensure all running services begin writing to the new locaiton. TcpTimedWaitDelay Defines the time that must elapse before TCP can release a closed connection and reuse its resources. For more information, refer to this Microsoft Article. Key Location: HKLM\SYSTEM\CurrentControlSet\Services\Tcpip\Parameters\Value Name: TcpTimedWaitDelayValue Type: DWORD (32-Bit) ValueDefault Value Data (Dec): 240Recommended Value Data (Dec): 30 Note: Machine reboot required to apply changes.
If unsure whether an advanced registry setting (even those not listed in this article) applies to your situation, please get in touch with Veeam Technical Support for assistance.