This article provides information about the impact of Automatic Maintenance Tasks on overall server performance, on a server pair protected by Neverfail.
Automated Maintenance Tasks include database or index defragmentation and compression, backups, file-level antivirus scans etc., which are launched on the server according to a schedule. Such tasks can result in a significant increase in processor/disk activity and may be accompanied by the generation of log files. As a result of the increase in replication traffic, performance problems may arise (mostly in WAN implementations).
In some situations, updates can accumulate in the send (unsafe) queue. This backlog can impact application performance, in which case clients may experience poor response times as the application slows down. This may happen if the rate at which data is generated on the active server exceeds the available bandwidth, or if the rate at which data is generated on the active server exceeds the rate at which the passive server can apply the changes.
These situations can be managed by adjusting the maintenance schedule so that the effect of any backlog is minimized and confined to the off-peak hours.
Neverfail recommends the following:
- Use the Neverfail SCOPE report to estimate if (and when) the measured bandwidth is adequate for performing maintenance tasks.
- Adjust the maintenance period according to findings in 1), so that the backlog has time to clear before the peak hours.
- Ensure that there is not more than one maintenance task running at any time and allow at least 30 minutes between two consecutive tasks.
- Avoid performing a Full System Check on the protected file system during maintenance periods.