Troubleshooting and FAQ

Find the container name#

The commands on this page use the name of the Bekci container. It depends on how you installed:

docker ps --filter ancestor=kadirsungurlu/bekci:latest

(If you use a fixed version, write that tag instead of latest, e.g. kadirsungurlu/bekci:1.0.0.)

I forgot my password#

If you can't sign in, you can reset the password from the server. Run this while Bekci is running, with your own username instead of admin:

cd /opt/bekci
docker compose exec bekci uptime sifre-sifirla admin

docker exec -it bekci uptime sifre-sifirla admin

uptime sifre-sifirla admin

(sifre-sifirla is Turkish for “reset password”.) The command asks Yeni şifre: (“New password:”); type the new password (at least 8 characters) and press Enter. The password is visible while you type it, so mind who's watching.

To turn off two-factor authentication as well (if you lost both your phone and your recovery codes), add --2fa-kapat at the end:

docker compose exec bekci uptime sifre-sifirla admin --2fa-kapat

The command works the same way with both images (SQLite and PostgreSQL).

Is Bekci running?#

The health check address is /healthz; it returns ok when the database is reachable:

curl http://localhost:8080/healthz          # from inside the server (if the port is published)
curl https://bekci.example.com/healthz          # from outside

The STATUS column of docker ps also shows the image's own health check: (healthy) means all is well, (unhealthy) means there's a problem. If the database can't be reached, /healthz returns 503 with veritabanı erişilemiyor (“database unreachable”).

Logs#

docker logs --tail 100 bekci                 # docker run
cd /opt/bekci && docker compose logs --tail 100 bekci   # Compose

On Coolify, look at the application's Logs. For more detail set LOG_LEVEL=debug and recreate the container (environment variables). The logs are written in Turkish; after a failed start the last line begins with hata: (“error:”) and gives the reason.

The container waits, saying the data is “open in another instance”#

If you see this in the log:

level=WARN msg="veri klasörü başka bir örnekte açık; o kapanana kadar bekleniyor" kilit=/data/uptime.db.lock en_fazla=10m0s
level=WARN msg="hâlâ bekleniyor: diğer örnek veri klasörünü bırakmadı" gecen=30s

(“the data folder is open in another instance; waiting until it closes” / “still waiting: the other instance hasn't released the data folder”), a second Bekci container was started with the same data volume. So that two copies never run checks and send notifications twice, the new copy waits until the old one stops.

• During an update on Coolify this is normal: once the old container stops, the new one takes over within a second or two.
• In any other situation, find the old container using the same volume and stop it: docker ps -a. If the lock isn't released within 10 minutes (UPTIME_LOCK_WAIT), the new container exits with the error aynı veri klasörüyle iki uygulama çalıştırılamaz (“two instances can't run with the same data folder”).

Port already in use#

A port is already allocated or address already in use error means another program uses the port you chose:

docker ps --filter publish=8080        # the container using the port
ss -ltnp 'sport = :8080'              # the program using the port

• 8080: on servers with Coolify, Coolify's proxy uses 8080. Pick another port (-p 8081:8080) or follow the Coolify guide.
• 80 / 443: a web server already runs on the server. Instead of the Caddy install, follow the reverse proxy or Coolify guide.

The panel opens but live updates don't arrive#

If the Offline notice keeps showing at the top of the page, or monitor states only change when you reload, the reverse proxy in front of Bekci is buffering the live stream (/api/events). The setting your proxy needs is on the reverse proxy page.

“Too many failed attempts”#

After 5 failed sign-in attempts from the same IP address within 15 minutes, that address has to wait 15 minutes. Try again when the time is up, or reset the password. If Bekci sits behind a reverse proxy and all users get blocked at once, the proxy isn't passing the X-Forwarded-For header (reverse proxy settings).

Times look wrong#

The interface currently always shows times in Türkiye time (Europe/Istanbul). Daily summaries, the nightly backup time and the log lines follow the server's TZ environment variable (default Europe/Istanbul). For another time zone set e.g. TZ=Europe/Berlin and recreate the container. The Bekci başladı (“Bekci started”) log line shows the time zone in use (saat_dilimi=).

Links in notifications or the address in agent commands are wrong#

BASE_URL isn't set or is wrong. Set it to the address the panel is opened at from outside (https://…) and recreate the container (environment variables).

The disk is full#

When the disk is full, Bekci can't write new measurements; with SQLite the automatic backup before a database-upgrading update can't be taken either, and Bekci won't start. Free up space: old pre-update backups (/data/backups/pre-migrate-…db) are never deleted automatically, and unused Docker images take up space too. You can shorten how long raw check history is kept under Settings → General.

PostgreSQL version mismatch#

If the postgres image refuses to start with HATA: PostgreSQL ana sürümü uyuşmuyor. (“ERROR: PostgreSQL major version mismatch.”), follow the PostgreSQL major version upgrade steps.

Frequently asked questions#

Is Bekci free?#

Yes. Bekci is open source and released under the AGPL-3.0 license; no license fees, no subscription, no paid tiers. The source code is on GitHub.

Does it need an internet connection?#

The image is downloaded from Docker Hub during install and updates. After that Bekci only connects out for the checks and notifications you define; it can also watch targets on your internal network.

Can I run two copies of Bekci at the same time?#

Not with the same data; the second copy waits until the first one stops. To check from different locations, use a check location.

Can I move from SQLite to PostgreSQL?#

The database can't be moved directly. You can carry monitors, notification channels, tags and status pages over to a new install with the JSON backup under Settings → Backup / Restore; users and check history aren't included (details).

How much disk does it use?#

It depends on the number of monitors and the check interval. Each individual check is kept for 14 days by default, then condensed into hourly and daily summaries; you can change these periods under Settings → General. Nightly backups take a few times the database size as well.

I found a bug or want a feature#

Open an issue on GitHub or write to [email protected].