Performance and Throttling Dialog
■ Bandwidth Throttling: The maximum bandwidth that Zerto Virtual Replication uses from this site to recovery sites.
The default value is for Zerto Virtual Replication to automatically assign the bandwidth used per VPG, based on using the maximum available and then prioritizing the usage according to the priority set for the VPGs sending data over the WAN. The minimum supported bandwidth is 5 Mb/sec.
■ Select Limited to define the limit. The default bandwidth throttling is Limited.
■ In the text box, set the Mb/sec. The valid range is from 0 to 100 Mb/sec.
With 0 Mb/Sec, Zerto Virtual Replication automatically assigns the bandwidth used per VPG, based on using the maximum available and then prioritizing the usage according to the priority set for the VPGs sending data over the WAN.
Enter the Mb/sec. when the value required is 100 Mb/sec. or more.
■ Time-based Throttling: To throttle the bandwidth during specific times. For example, during the daily peak transaction period you can change the bandwidth throttling, to override the general setting.
■ Limited: Select to define the limit. The default time-based throttling is Limited.
■ From: The hour and the minute to start the throttling, using a 24-hour clock.
■ To: The hour and the minute to end the throttling, using a 24-hour clock.
IMPORTANT: |
Performance and Throttling > advanced settings must only be changed in coordination with Zerto support. |
■ Enable Bandwidth Regulation: To enable regulating the bandwidth to troubleshoot problems.
■ Enable IO throttling: If a host is handling too many IOs, then the IOs begin to get high latencies. To offset this the VRA sends fewer concurrent IOs. The latency is measured by taking the average latency for all IOs over a set period of time. For example, when the period is 5000 milliseconds and the bad IO latency is 40, the average latency is calculated every 5 seconds, and if the average latency exceeds 40, the VRA sends fewer concurrent IOs.
■ Bad IO Latency VM: The threshold above which the latency is considered high, and therefore bad.
■ Requested Duration (ms): The period of time used to measure the average latency.
