Nothing. I always use the standard configuration from the services I’ve installed. If something doesn’t work I go into those logs and look what’s wrong.

Honestly? I just ignore them. Something seems to be happening to them, as they’re not growing infinitely, but no idea why.

It was a bit complicated to set this up, but it took me almost no time at all.

I’ve just started digging into it myself ! Here’s my current setup (I’ll see how it scales in the long term):

• syslog on every host
• Telegraf collects and parse logs
• InfluxDB stores everything
• Grafana for dashboards

I run OpenBSD on all my servers, and configure all the services to log via syslog.

Then I configuré syslog to send only those I care about (https, DNS, …) to a central telegraf instance, using the syslog protocol (RFC3164).

On this collector, telegraf gets all these logs and parse them using custom grok patterns I’m currently building, to make sense out of every log line it receives. The parsed logs are in turns stored in Influxdb, running on the same host.

I then use Grafana to query InfluxDB and create dashboards out of these logs. Grafana can also display the logs “as-is” so you can search through them (it’s not ideal though as you simply search by regex from the full message, so it’s on par with grep at least).

This setup is fairly new and seem to work very well. Telegraf is also very low on resource usage for now. I’ll have to continue adding grok patterns and send more application logs to it to see how it handles the load. I do have a few questions still unanswered for now, but time will tell:

Q: Should I first collect via a central syslog before sending to telegraf ?
This would let syslog archive all logs in plain text, rotate and compress them. I would also only have a single host to configure for sending logs to telegraf. However this would eat up space, and could hide the original sending hostname for each log. I might try that someday.

Q: Should I run telegraf on each host ?
This would distribute the load of the grok parsing amongst all hosts, and then all telegraf processes will send directly to the central one for collection, or even directly into influxdb. I would also benefit from telegraf being install on each host to collect more data (CPU, network stats, …). However it makes the configuration more complex to handle.

Q: What is a good retention period ?
For now, influxDB doesn’t expire any data, as I don’t have much yet. In the long run, I should probably delete old data, but it’s hard to tell what is “old” in my case.

Q: Do I need an interface to read logs ?
I use this setup mostly for graphs, as grafana can make sense of fields like “http_verb”, “http_code” and such. However, it is much more practical for me to dig into the logs right on the server, in /var/log. Having an interface like chronograf or graylog seems practical, but I feel like it’s overdoing it.

Bonus:

For a bit enhanced log file viewing, you could use something like lnav, I think it’s packaged for most distributions.

Cockpit can be useful for journald, but personally I think GUI stuff is a bit clunky for logs.

Grep, awk and sed are powerful tools, even with only basic knowledge of them. Vim in readonly mode is actually quite effective for single files too.

For aggregating multiple servers’ logs good ol’ rsyslog is good, but not simple to set up. There are tutorials online.

Personally I use Loki to aggregate my logs but I’m probably not the typical selfhoster.

You should first define your needs, if you don’t need logs, then don’t bother with them.

Syslog (rsyslod) is usually the standard answer for the average sysadmin, but it depends a lot on your needs. A lot of newer loggers output as pure JSON, which offer benefits to readability and more approachable search logic/filters/queries (I’m so tired of regex).

When you start venturing down the road of finding the right way to store and forward the output of logging drivers from Docker containers, as one example, rsyslod starts to feel dated.

The easy answers if you want to throw money at the problem are solutions like Splunk, Datadog, or New Relic. If you don’t want to (and most people wouldn’t), then alternatives certainly exist, but some of them are just as heavy on system resources. Greylog has relative feature parity with Splunk Enterprise, but consumes just as much compute and storage if not more, and I found it to be a much larger pain in the butt to administer and keep running.

The likeliest answer to this problem is Grafana Loki, just based on what I’ve read of its capabilities, but I haven’t had a chance to circle back and test it out. Someone here who has might be able to weigh in and speak to its strengths/weaknesses.

In terms if logging: sys journal does the job for me …

I do run prometheus + grafana for some services but that is mostly for some fancy looking graphics nothing really usefull.

I would recomend you to monitor updates of you apps so you are well informed when und what to update (i just have subsribed to all the diffrent release git rss feeds)

My logs were actually filling up my disk! I edited journald conf and limited them to 500mb . disk is only 10gb for reference

You could try deploying an ELK stack and manage log collection using Elastic Agents and integrations, works pretty well for us and, as long as your log is supported by an integration, it’s dead simple to manage

Systemd does all that for you. Just set a limit in journal.conf

I use the standard system syslog with logrotate every 7 days with 1 month of gzip archive.

However on production systems, I run a central rsyslog server which archives once a week and a year of archives. Considering ELK in the future but for simple retention syslog is fine.

I’ve been using Loki and promtail for containers I actually care about aggregating logs for. It’s something relatively new for me but so far the experience has been positive.

I’ve done everything from local syslog to various open source and proprietary log pipelines and also worked on a team dedicated to building a custom log-processing pipeline from scratch in a large environment.

One lesson I’ve learned is that no matter how nice that web UI is I still want some kind of CLI access and the equivalent of being able to cat, tail and grep logs. In large environments it won’t be actually tail-ing a logfile but I want something that can produce a live logstream for piping into Unix text (and JSON) processing tools.

Make sure that you have logrotate installed and working, to prevent them from piling up.

Greylog