First signs that Clustermode collectors are reaching their limits in larger environments are emerging. In their current form, the collectors are not able to finish collecting and storing all the data within 60 seconds. Therefore, we have optimized the collectors for volume and col_snapshot for experimental purposes. We managed to reduce the size of collected data by a factor of 4-5. Thus, the filer doesn’t have to provide and deliver this data. This helps reduce the workload on filers caused by the monitoring operation.
As we know, collectors provide data for the monitoring process in a very cost effective way in order to minimize the frequent collection of the same data.
The performance data collector (
get_netapp_perfdata) has already been fully optimized. Should this result in timeouts, please forward them by means of the
‑‑mammamia report (available in version 3.0.15 or higher). This way we can make sure if we can improve our checks or if its a problem on the filer. In the latter case, we can provide input by means of opening a Case at NetApp.