We should evaluate the migration of Matomo to a new monitoring server. Also because of the current server intreccio is under visible stress.
But, we can also opt to stretch up intreccio as well, of course, right now.
Description
Description
| Status | Subtype | Assigned | Task | ||
|---|---|---|---|---|---|
| Open | None | T338175 WordPress performance and downtime issues | |||
| Open | stefano.cannillo | T338353 Evaluate the migration of Matomo to a new monitoring server | |||
| Open | stefano.cannillo | T338350 Evaluate a new server in a potentially different service provider to migrate in it Matomo and also setup a Zabbix monitoring system | |||
| Resolved | valerio.bozzolan | T338358 Grant to Stefano Cannillo access on the "vh" service provider |
Event Timeline
Comment Actions
It's good to avoid putting all eggs in one basket, but how did you determine that Matomo is the culprit? A more obvious first step would be to decouple MediaWiki and Wordpress which are probably the two largest applications (plus their load is heavily correlated and it would be great if they didn't go down at the same time, while Matomo's load is mostly in batch jobs which get be scheduled off-peak).
When you migrate Matomo away, it would be a good chance to consider using a managed instance, to reduce maintenance work. One option is https://cloud68.co/instances/matomo , as WMI already uses Cloud68 (whose data is accessible to employees outside the EU jurisdiction and outside states with adequacy decisions, but not in USA). Otherwise there's Matomo Cloud by Innocraft (which is in EU, New Zealand, Canada and Japan, all states with adequacy decisions). Though if it were my personal website I'd probably go with Plausible.