# Confused About the Config File

**URL:** <https://kopia.discourse.group/t/confused-about-the-config-file/9014>\
**Category:** General\
**Created:** [September 17, 2026, 3:40pm UTC](https://kopia.discourse.group/t/confused-about-the-config-file/9014 "2026-09-17T15:40:09Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![bpc](https://avatars.discourse-cdn.com/v4/letter/b/edb3f5/32.png) [@bpc](https://kopia.discourse.group/u/bpc)\
**Post date:** [September 17, 2026, 3:40pm UTC](https://kopia.discourse.group/t/confused-about-the-config-file/9014/1 "2026-09-17T15:40:09Z")

</div>

Howdy! I’m pretty new to Kopia, but like what I see. I’ve been wanting to ditch Duplicati for a while & and finally doing it. I was attracted to Kopia for The GUI, but have so far found no use for it. I’m running kopia CLI from cron on a couple Linux systems, including my daily driver laptop.

I spent quite some time editing the config files before discovering that they are ephemeral. They appear to be created and destroyed with every connect/disconnect. Editing them seems to do nothing, yet I see them discussed in the forum and there are flags on the CLI commands to choose config files. I’m very confused by their purpose and whether I should care about them. It seems that Repository configuration takes place in Policy. Can some kind soul explain config files to me?

---

<div class="post-metadata">

**Author:** ![sfatula](https://avatars.discourse-cdn.com/v4/letter/s/898d66/32.png) [@sfatula](https://kopia.discourse.group/u/sfatula)\
**Post date:** [September 20, 2026, 11:40pm UTC](https://kopia.discourse.group/t/confused-about-the-config-file/9014/2 "2026-09-20T23:40:14Z")

</div>

If you are using the CLI, then, include the config file options in the command line for each command that offers it. Otherwise, they do nothing most likely as you say. The “connection” is not a connection really, it doesn’t maintain an open connection like you may be thinking.

---

<div class="post-metadata">

**Author:** ![phunsoft](https://avatars.discourse-cdn.com/v4/letter/p/c2a13f/32.png) [@phunsoft](https://kopia.discourse.group/u/phunsoft)\
**Post date:** [September 22, 2026, 7:51pm UTC](https://kopia.discourse.group/t/confused-about-the-config-file/9014/3 "2026-09-22T19:51:09Z")

</div>

When connecting to a repository using CLI, this “connection” is store in a “repository.config” file. This defines the default repository (aka connection) used by the CLI when you don’t specify another config file using “–config-file”.

When you disconnect the repository using the CLI, then this file is removed.

So, the CLI remembers only a single connection (unless you specify “–config-file”)

The Kopia UI, on the other hand, can remember multiple connections. It does so by appending a UNIX timestamp to the config file name, e.g. “repository-1789739686572.config” when connecting to a currently not connected repository.

This is quite handy,at least in my case. I’m using a repository on each of two USB HDDs, each on with a unique pathname to the repository (e.g. “backupdisk-A/repositoyA”, and “backupdisk-B/repositoyB”). They contain the same snapshots. I regularly swap the HDDs, and store one in another location. Kopia UI remembers both repositories and has one config file for each one. No matter which HDD I connect to the PC, Kopia UI automatically connects to the repository on the HDD. This even works for both HDDs connected in parallel. I can choose which one to work with from the Koipa UI task bar icon. I really like this.

If there is a “CLI” config file, i.e. one without the UNIX timestamp in the name, the Kopia UI shows this also in the list of connected repositories.

If a repository (in my case, one or more of the HDDs) is not connected, the Kopia UI shows “initializing”. Reconnect the device and Kopia UI recognized the repository to be available automatically.