mirror of
https://github.com/mailscope/kumomta.git
synced 2026-08-23 12:48:17 +00:00
Wez Furlong
112 lines
4.3 KiB
Markdown
112 lines
4.3 KiB
Markdown
# Getting Server Status
|
|
|
|
Once KumoMTA is installed, you can check on the server status with systemctl.
|
|
|
|
```console
|
|
$ sudo systemctl status kumomta
|
|
```
|
|
|
|
The result should look something like this:
|
|
|
|
```
|
|
kumomta.service - KumoMTA SMTP service
|
|
Loaded: loaded (/lib/systemd/system/kumomta.service; enabled; vendor preset: enabled)
|
|
Active: active (running) since Thu 2023-04-27 22:59:06 MST; 10h ago
|
|
Main PID: 17912 (kumod)
|
|
Tasks: 28 (limit: 19190)
|
|
Memory: 257.1M
|
|
CGroup: /system.slice/kumomta.service
|
|
└─17912 /opt/kumomta/sbin/kumod --policy /opt/kumomta/etc/policy/init.lua --user kumod
|
|
|
|
Apr 27 22:59:06 kdev2.kumomta.com systemd[1]: Started KumoMTA SMTP service.
|
|
Apr 27 22:59:06 kdev2.kumomta.com kumod[17912]: 2023-04-28T05:59:06.444479Z INFO main kumod::memory: using limits: soft=Some("12.58 GB"), hard=So>
|
|
Apr 27 22:59:06 kdev2.kumomta.com kumod[17912]: 2023-04-28T05:59:06.450824Z INFO localset-2 kumod::http_server: http listener on 0.0.0.0:8000
|
|
Apr 27 22:59:06 kdev2.kumomta.com kumod[17912]: 2023-04-28T05:59:06.471926Z INFO localset-2 kumod::spool: starting enumeration for meta
|
|
Apr 27 22:59:06 kdev2.kumomta.com kumod[17912]: 2023-04-28T05:59:06.471995Z INFO localset-2 kumod::smtp_server: smtp listener on 0.0.0.0:25
|
|
Apr 27 22:59:06 kdev2.kumomta.com kumod[17912]: 2023-04-28T05:59:06.472008Z INFO localset-2 kumod::smtp_server: smtp listener on 0.0.0.0:2026
|
|
Apr 27 22:59:06 kdev2.kumomta.com kumod[17912]: 2023-04-28T05:59:06.475882Z INFO localset-2 kumod: initialization complete
|
|
```
|
|
|
|
The above is from a newer installation, but the log will grow. If you are
|
|
debugging an older install, `journalctl -r -n 10 -u kumomta.service` will show
|
|
the last 10 lines (`-n 10`) in reverse order (`-r`). `man journalctl` is your
|
|
friend.
|
|
|
|
If you need to find the installed version, you can run:
|
|
|
|
```console
|
|
$ /opt/kumomta/sbin/kumod --version
|
|
```
|
|
|
|
This will be important if you ever need to reach out for support.
|
|
|
|
## Monitoring
|
|
|
|
If you have configured an HTTP listener, you can access server metrics in Prometheus format with:
|
|
|
|
```console
|
|
$ curl -i 'http://localhost:8000/metrics'
|
|
```
|
|
|
|
That will show a long form of the server metrics with detailed comments.
|
|
|
|
If you want just the data in a nice JSON format, use:
|
|
|
|
```console
|
|
$ curl -i 'http://localhost:8000/metrics.json'
|
|
```
|
|
|
|
Metrics available include the following at the time of writing, and will
|
|
increase as we build out the product:
|
|
|
|
* `connection_count`: number of active connections
|
|
* `lua_count`: the number of lua contexts currently alive
|
|
* `lua_load_count`: how many times the policy lua script has been loaded into a new context
|
|
* `lua_spare_count`: the number of lua contexts available for reuse in the pool
|
|
* `memory_limit`: soft memory limit measured in bytes
|
|
* `memory_usage`: number of bytes of used memory
|
|
|
|
## Using kcli
|
|
|
|
Two handy commands are:
|
|
|
|
* [kcli queue-summary](../../reference/kcli/queue-summary.md) to show a textual representation of queue depths
|
|
* [kcli top](../../reference/kcli/top.md) to show a TUI charting top-line metrics
|
|
|
|
These depend upon having the HTTP listener configured so that the metrics
|
|
endpoint can be accessed.
|
|
|
|
## Setting up a Grafana Dashboard
|
|
|
|
You will need to [install
|
|
Prometheus](https://prometheus.io/docs/prometheus/latest/installation/) to act
|
|
as a datasource, and [install
|
|
Grafana](https://grafana.com/docs/grafana/latest/setup-grafana/installation/)
|
|
for its dashboard capabilities.
|
|
|
|
With Prometheus installed and available, configure it to collect data from
|
|
your kumod instances using a scraper configuration like this:
|
|
|
|
```yaml
|
|
scrape_configs:
|
|
- job_name: kumomta
|
|
scrape_interval: 5s
|
|
metrics_path: /metrics
|
|
static_configs:
|
|
- targets:
|
|
- 'kumomta-1:8000'
|
|
- 'kumomta-2:8000'
|
|
```
|
|
|
|
In the above configuration `kumomta-1:8000` is the hostname (or IP address) of one
|
|
of the machines running kumod, and `8000` is the HTTP listener port it has open.
|
|
`kumomta-2:8000` is the `IP:port` of another instance of kumod; you can list out
|
|
as many as are present in your infrastructure.
|
|
|
|
With the Prometheus datasource in place, you can [import our starter
|
|
dashboard](https://grafana.com/grafana/dashboards/21391-kumomta/) in the
|
|
Grafana UI by entering its ID number `21391` or otherwise downloading the JSON
|
|
[from the dashboard
|
|
page](https://grafana.com/grafana/dashboards/21391-kumomta/) and then importing
|
|
it into your Grafana instance directly.
|