Long-running tasks

There are a few tasks Shlink needs to perform which are too heavy or time-consuming to be executed when a user visits a short URL.

In order to redirect end users as fast as possible to the final URLs, these tasks have been left for you to configure when and how often should they be executed.

Note

These tasks are not strictly required when serving Shlink with openswoole or RoadRunner, as they have their own background jobs system that Shlink uses to delegate these tasks.

Download GeoLite2 DB

The visit:download-db command can be used to make sure your GeoLite2 db is up-to-date.

GeoLite2 DBs do not change that frequently or that much, so you don’t need to schedule this one with a very high frequency.

Locate visits

Shlink provides the command visit:locate, which will process all the IP addresses from visits in order to locate them. Those that have already been located will be skipped.

Shlink locates visits automatically just before redirecting end users. However, it might be a good idea to have previous scheduling in place, as a backup in case it fails.

Note

When using the docker image and Shlink v2.8.0 or newer, you can provide -e ENABLE_PERIODIC_VISIT_LOCATE=true in order to schedule visits to be located every hour inside the container itself.

Using cronjobs

One way to schedule any of the previous tasks is by making use of Unix cron jobs:

  • Locate visits every hour: 0 * * * * /path/to/shlink/bin/cli visit:locate -q -n
  • Update the GeoLite2 db every 7 days: 0 0 * * 0 /path/to/shlink/bin/cli visit:download-db -q -n

The -q and -n flags are available on any Shlink command. The first one will discard any output, and the second one ensure the command is run in a non-interactive mode.