47 lines
3.3 KiB
Markdown
47 lines
3.3 KiB
Markdown
# Dashboards
|
|
|
|
Pleroma comes with two types of backend dashboards viewable to instance administrators:
|
|
|
|
* [Phoenix LiveDashboard](https://hexdocs.pm/phoenix_live_dashboard/Phoenix.LiveDashboard.html) - A general system oriented dashboard for viewing statistics about Pleroma resource consumption, Pleroma's database and Pleroma's job processor (Oban).
|
|
* [Oban Web](https://hexdocs.pm/oban_web/overview.html) - A dashboard specific to Oban for viewing Oban statistics, managing jobs and job queues.
|
|
|
|
!!! note
|
|
Both dashboards require working Websockets.
|
|
If your browser or web server don't support Websockets, both dashboards either won't update or will not display all information.
|
|
|
|
## Phoenix LiveDashboard
|
|
|
|
Instance administrators can access this dashboard at `/pleroma/live_dashboard`, giving a simple overview of software versions including Erlang and Elixir versions, instance uptime and resource consumption.
|
|
|
|
This dashboard gives insights into the current state of the BEAM VM running Pleroma code and database statistics including basic diagnostics.
|
|
It can be useful for troubleshooting of some issues namely regarding database performance.
|
|
|
|
### Relevant dashboard tabs
|
|
|
|
* Home - A general overview of system information including software versions, uptime and memory BEAM memory consumption.
|
|
* OS Data - Information about the OS and system such as CPU load, memory usage and disk usage.
|
|
* Ecto Stats - Information about the Pleroma database.
|
|
- Diagnose - Basic database diagnostics, including a `bloat` warning when an index or a table have excessive bloat, which can lead to bad database performance.
|
|
- Bloat - A table showing size of "bloat" (unused wasted space) in database tables and indexes. Very high bloat size in the `activities` and `objects` tables can lead to bad performance especially on slower disks such as on most VPS providers.
|
|
- Db settings - A small list of PostgreSQL settings mostly relevant to database performance.
|
|
- Total table size - Shows sizes of all database tables including indexes sorted by size, useful for quickly checking overall database size.
|
|
- Long running queries - A list of of slow database queries and their duration. Multiple entries with duration in multiple seconds indicate a slowly performing database.
|
|
* Oban - Shows a list of all Oban jobs.
|
|
|
|
!!! note
|
|
The DB bloat warning for `index 'oban_jobs::oban_jobs_args_index'` in Ecto Stats can be safely ignored.
|
|
|
|
## Oban Web
|
|
|
|
An advanced dashboard and management console viewable to instance administrators specifically for Oban, Pleroma's job processor.
|
|
It allows managing jobs, including force retrying failed jobs and job deletion.
|
|
It can be accessed at `/pleroma/oban`.
|
|
|
|
!!! danger
|
|
This dashboard is very powerful! If you are unsure what a certain feature does, don't use it.
|
|
Changing individual queue state/settings in the "Queues" view is heavily discouraged.
|
|
|
|
* Shows a real time chart of either a number of executed jobs, or job execution/wait time per a given time frame and the state/queue/worker.
|
|
* Shows a list of jobs in each state, their argument, number of attempts and execution/scheduled time.
|
|
* Selecting one or multiple jobs in the list allows performing actions like canceling/deleting and retrying.
|
|
* Clicking on a job shows a detailed view including the full argument, when it was inserted, information about its attempts, and performing actions on it.
|