https://prefect.io logo
Join Slack
Powered by
<@ULVA73B9P> i have to run my prefect server on po...
# ask-marvin
s
@Marvin i have to run my prefect server on postgresql with authstrings (.venv) root@Prefect:/home/prefect# prefect version Version: 3.6.15 API version: 0.8.4 Python version: 3.12.3 Git commit: 757a089b Built: Fri, Jan 30, 2026 12:27 AM OS/Arch: linux/x86_64 Profile: server Server type: server Pydantic version: 2.12.5 Server: Database: sqlite SQLite version: 3.45.1 Integrations: prefect-docker: 0.7.1 (.venv) root@Prefect:/home/prefect# prefect config view 🚀 you are connected to: 192.168.29.151 PREFECT_PROFILE='server' PREFECT_API_AUTH_STRING='********' (from profile) PREFECT_API_URL='192.168.29.151/api' (from profile) (.venv) root@Prefect:/home/prefect# cat /etc/systemd/system/prefect-server.service [Unit] Description=Prefect Server After=network-online.target postgresql.service Wants=network-online.target [Service] Type=simple User=root WorkingDirectory=/home/prefect Environment="PREFECT_HOME=/home/prefect/.prefect" Environment="PREFECT_SERVER_API_HOST=192.168.29.151" Environment="PREFECT_SERVER_API_PORT=4200" Environment="PREFECT_SERVER_API_KEEPALIVE_TIMEOUT=60" Environment="PREFECT_SERVER_LOGGING_LEVEL=DEBUG" Environment="PREFECT_API_DATABASE_CONNECTION_URL=postgresql+asyncpg://prefect:Network%405714@127.0.0.1:5432/prefect" Environment="PREFECT_SERVER_API_AUTH_STRING=astra:Network@5714" ExecStart=/home/prefect/.venv/bin/prefect server start --host 0.0.0.0 --port 4200 --keep-alive-timeout 60 Restart=on-failure RestartSec=10 StandardOutput=journal StandardError=journal [Install] WantedBy=multi-user.target (.venv) root@Prefect:/home/prefect# not sure it is using sqlite or postgresql
m
✅ thought for 121.4 seconds
Yes — based on what you pasted, Prefect is still using SQLite right now. Why I say that: - 
prefect version
shows: - 
Server:
- 
Database: sqlite
- 
prefect config view
does not show any database connection setting at all - In your systemd unit, the PostgreSQL URL is set as: - 
PREFECT_API_DATABASE_CONNECTION_URL=...
- But your current shell session where you ran 
prefect version
/ 
prefect config view
does not appear to have that env var loaded A couple important points: 1. 
PREFECT_API_DATABASE_CONNECTION_URL
is the correct setting Prefect 3 supports that variable for the server DB connection. Your value should look like:
Copy code
bash
<postgresql+asyncpg://prefect:Network%405714@127.0.0.1:5432/prefect>
2. Your systemd file has a mistake You currently have angle brackets around the connection string:
Copy code
ini
Environment="PREFECT_API_DATABASE_CONNECTION_URL=<postgresql+asyncpg://prefect:Network%405714@127.0.0.1:5432/prefect>"
Those 
< >
should not be there. Use:
Copy code
ini
Environment="PREFECT_API_DATABASE_CONNECTION_URL=<postgresql+asyncpg://prefect:Network%405714@127.0.0.1:5432/prefect>"
Also, since your password contains 
@
, URL-encoding it as 
%40
is correct. 3. How to verify what the running systemd service is using Run these:
Copy code
bash
systemctl daemon-reload
systemctl restart prefect-server
systemctl show prefect-server -p Environment
journalctl -u prefect-server -n 100 --no-pager
That will show whether the service actually has the DB env var. Also check from the same shell as the service config expectation:
Copy code
bash
prefect config view --show-secrets | grep DATABASE
prefect version
If Prefect is using PostgreSQL, 
prefect version
should show something like:
Copy code
text
Server:
  Database: postgresql
  PostgreSQL version: ...
4. Recommended corrected service file
Copy code
ini
[Unit]
Description=Prefect Server
After=network-online.target postgresql.service
Wants=network-online.target

[Service]
Type=simple
User=root
WorkingDirectory=/home/prefect
Environment="PREFECT_HOME=/home/prefect/.prefect"
Environment="PREFECT_SERVER_API_HOST=192.168.29.151"
Environment="PREFECT_SERVER_API_PORT=4200"
Environment="PREFECT_SERVER_API_KEEPALIVE_TIMEOUT=60"
Environment="PREFECT_SERVER_LOGGING_LEVEL=DEBUG"
Environment="PREFECT_API_DATABASE_CONNECTION_URL=<postgresql+asyncpg://prefect:Network%405714@127.0.0.1:5432/prefect>"
Environment="PREFECT_SERVER_API_AUTH_STRING=astra:Network@5714"

ExecStart=/home/prefect/.venv/bin/prefect server start --host 0.0.0.0 --port 4200 --keep-alive-timeout 60

Restart=on-failure
RestartSec=10
StandardOutput=journal
StandardError=journal

[Install]
WantedBy=multi-user.target
Then:
Copy code
bash
systemctl daemon-reload
systemctl restart prefect-server
journalctl -u prefect-server -f
5. Best quick test Temporarily export the DB URL in your current shell and run:
Copy code
bash
export PREFECT_API_DATABASE_CONNECTION_URL="<postgresql+asyncpg://prefect:Network%405714@127.0.0.1:5432/prefect>"
prefect config view --show-secrets | grep DATABASE
prefect version
If that switches 
prefect version
from 
sqlite
to 
postgresql
, then the issue is just your systemd environment/config. Docs / references - Prefect settings and configuration: Prefect settings - Prefect self-hosted server docs: Prefect Server If you want, I can also help you with a step-by-step PostgreSQL migration checklist for this exact server.
s
@Marvin this line had issue Environment="PREFECT_API_DATABASE_CONNECTION_URL=postgresql+asyncpg://prefect:Network%405714@127.0.0.1:5432/prefect" there is no <> in url in your chat on past it looks like
m
✅ thought for 4.8 seconds
Yes — exactly. The problem is the literal 
<
and 
>
characters in the systemd file. This is wrong:
Copy code
ini
Environment="PREFECT_API_DATABASE_CONNECTION_URL=<postgresql+asyncpg://prefect:Network%405714@127.0.0.1:5432/prefect>"
This is correct:
Copy code
ini
Environment="PREFECT_API_DATABASE_CONNECTION_URL=<postgresql+asyncpg://prefect:Network%405714@127.0.0.1:5432/prefect>"
I used 
<...>
earlier as a placeholder style in explanation, but in the actual config file you must not include them. So your fixed line should be:
Copy code
ini
Environment="PREFECT_API_DATABASE_CONNECTION_URL=<postgresql+asyncpg://prefect:Network%405714@127.0.0.1:5432/prefect>"
Then run:
Copy code
bash
systemctl daemon-reload
systemctl restart prefect-server
journalctl -u prefect-server -n 100 --no-pager
prefect version
If it worked, 
prefect version
should change from:
Copy code
text
Database: sqlite
to something like:
Copy code
text
Database: postgresql
If you want, paste the output of these and I’ll verify it:
Copy code
bash
systemctl show prefect-server -p Environment
prefect config view --show-secrets | grep DATABASE
prefect version
s
@Marvin (.venv) root@Prefect:/home/prefect# systemctl status prefect-server ● prefect-server.service - Prefect Server Loaded: loaded (/etc/systemd/system/prefect-server.service; enabled; preset: enabled) Active: active (running) since Wed 2026-07-08 201937 UTC; 6s ago Main PID: 3560 (prefect) Tasks: 1 (limit: 9431) Memory: 156.6M (peak: 156.7M) CPU: 6.930s CGroup: /system.slice/prefect-server.service └─3560 /home/prefect/.venv/bin/python3 /home/prefect/.venv/bin/prefect server start --host 0.0.0.0 --port 4200> Jul 08 201944 Prefect prefect[3560]: sqlalchemy.exc.ProgrammingError: (sqlalchemy.dialects.postgresql.asyncpg.Programming> Jul 08 201944 Prefect prefect[3560]: [SQL: Jul 08 201944 Prefect prefect[3560]: CREATE TABLE alembic_version ( Jul 08 201944 Prefect prefect[3560]: version_num VARCHAR(32) NOT NULL, Jul 08 201944 Prefect prefect[3560]: CONSTRAINT alembic_version_pkc PRIMARY KEY (version_num) Jul 08 201944 Prefect prefect[3560]: ) Jul 08 201944 Prefect prefect[3560]: ] Jul 08 201944 Prefect prefect[3560]: (Background on this error at: https://sqlalche.me/e/20/f405) Jul 08 201944 Prefect prefect[3560]: ERROR: Application startup failed. Exiting. Jul 08 201944 Prefect prefect[3560]: Server stopped!
m
✅ thought for 12.7 seconds
This means Prefect is now trying to use PostgreSQL, so that part is fixed. The new problem is a PostgreSQL permissions/schema issue during migration:
Copy code
text
sqlalchemy.exc.ProgrammingError
CREATE TABLE alembic_version ...
Prefect is attempting to create its migration table, but PostgreSQL is rejecting it. Most likely causes: - the 
prefect
user does not have permission to create tables in the target database - the database 
prefect
exists, but ownership/privileges are wrong - the connection is landing in a schema where the user lacks 
CREATE
permission What to check first Please run:
Copy code
bash
journalctl -u prefect-server -n 50 --no-pager
The important line is usually just above the SQL block and will say the exact PostgreSQL error, such as: - 
permission denied for schema public
- 
permission denied for database prefect
- 
must be owner of schema public
That exact message matters. Most common fix Connect to PostgreSQL as a superuser and make sure the DB/user ownership is correct.
Copy code
bash
sudo -u postgres psql
Then run:
Copy code
sql
ALTER DATABASE prefect OWNER TO prefect;
GRANT ALL PRIVILEGES ON DATABASE prefect TO prefect;

\c prefect

ALTER SCHEMA public OWNER TO prefect;
GRANT ALL ON SCHEMA public TO prefect;
GRANT ALL PRIVILEGES ON ALL TABLES IN SCHEMA public TO prefect;
GRANT ALL PRIVILEGES ON ALL SEQUENCES IN SCHEMA public TO prefect;
ALTER DEFAULT PRIVILEGES IN SCHEMA public GRANT ALL ON TABLES TO prefect;
ALTER DEFAULT PRIVILEGES IN SCHEMA public GRANT ALL ON SEQUENCES TO prefect;
If the DB/user were not created yet, a cleaner setup is:
Copy code
sql
CREATE USER prefect WITH PASSWORD 'Network@5714';
CREATE DATABASE prefect OWNER prefect;
\c prefect
GRANT ALL ON SCHEMA public TO prefect;
ALTER SCHEMA public OWNER TO prefect;
Then restart Prefect:
Copy code
bash
systemctl restart prefect-server
journalctl -u prefect-server -n 100 --no-pager
prefect version
Quick direct permission test You can test whether the user can create tables:
Copy code
bash
psql "<postgresql://prefect:Network%405714@127.0.0.1:5432/prefect>"
Then:
Copy code
sql
CREATE TABLE test_prefect_permissions(id int);
DROP TABLE test_prefect_permissions;
If 
CREATE TABLE
fails, that confirms it is a PostgreSQL privilege issue, not a Prefect issue. One more thing from your status output Your 
ExecStart
appears to show:
Copy code
text
prefect server start --host 0.0.0.0 --port 4200>
That trailing 
>
looks suspicious in the pasted output. If that is literally present in the service file, remove it so the command ends at:
Copy code
ini
ExecStart=/home/prefect/.venv/bin/prefect server start --host 0.0.0.0 --port 4200
Useful references - Prefect server docs: Prefect Server - Prefect settings docs: Prefect settings If you paste the full PostgreSQL error line from 
journalctl
, I can tell you the exact SQL fix.
s
@Marvin when we switch to postgresql database where to check the database disk usage
m
✅ thought for 9.5 seconds
If you're using PostgreSQL, check disk usage in PostgreSQL itself or on the server filesystem. Best PostgreSQL checks Connect first:
Copy code
bash
psql "<postgresql://prefect:Network%405714@127.0.0.1:5432/prefect>"
Then use these queries. 1. Size of the Prefect database
Copy code
sql
SELECT pg_size_pretty(pg_database_size('prefect')) AS db_size;
2. Size of all databases
Copy code
sql
SELECT datname,
       pg_size_pretty(pg_database_size(datname)) AS size
FROM pg_database
ORDER BY pg_database_size(datname) DESC;
3. Largest tables in the Prefect DB
Copy code
sql
SELECT
    schemaname,
    tablename,
    pg_size_pretty(pg_total_relation_size(schemaname || '.' || tablename)) AS total_size
FROM pg_tables
WHERE schemaname = 'public'
ORDER BY pg_total_relation_size(schemaname || '.' || tablename) DESC
LIMIT 20;
4. More detailed table breakdown
Copy code
sql
SELECT
    relname AS table_name,
    pg_size_pretty(pg_relation_size(C.oid)) AS table_size,
    pg_size_pretty(pg_indexes_size(C.oid)) AS indexes_size,
    pg_size_pretty(pg_total_relation_size(C.oid)) AS total_size
FROM pg_class C
LEFT JOIN pg_namespace N ON (N.oid = C.relnamespace)
WHERE nspname = 'public'
  AND relkind = 'r'
ORDER BY pg_total_relation_size(C.oid) DESC
LIMIT 20;
At the Linux filesystem level If PostgreSQL is local on the same host, you can also check where PostgreSQL stores data:
Copy code
bash
sudo -u postgres psql -t -c "show data_directory;"
Then check disk usage there:
Copy code
bash
du -sh /var/lib/postgresql
df -h
Or for the actual data directory path returned:
Copy code
bash
du -sh /path/to/postgresql/data_directory
If you only care about Prefect's old SQLite file Before migration, SQLite was usually under 
PREFECT_HOME
, likely:
Copy code
bash
/home/prefect/.prefect/prefect.db
Check it with:
Copy code
bash
du -sh /home/prefect/.prefect/prefect.db
Useful note After moving to PostgreSQL: - Prefect metadata is stored in the PostgreSQL database - flow/task logs may also contribute significantly depending on retention and logging volume - result storage is usually separate unless you've configured DB-backed result persistence somewhere custom If you want, I can give you: 1. a single SQL query for total Prefect DB usage, or 2. a query to find which Prefect tables grow the fastest.
s
@Marvin prefect service and ui is working as (.venv) root@Prefect:~/.prefect# systemctl show -p Environment --value prefect-server | tr ' ' '\n' PREFECT_HOME=/home/prefect/.prefect PREFECT_SERVER_API_HOST=192.168.29.151 PREFECT_SERVER_API_PORT=4200 PREFECT_SERVER_API_KEEPALIVE_TIMEOUT=60 PREFECT_SERVER_LOGGING_LEVEL=DEBUG PREFECT_API_DATABASE_CONNECTION_URL=postgresql+asyncpg://prefect:prefectpassword@127.0.0.1:5432/prefect PREFECT_SERVER_API_AUTH_STRING=astra:Network@5714 but why prefect config view at cli is still in different world are service and local cli in different worlds (.venv) root@Prefect:~/.prefect# systemctl show -p Environment --value prefect-server | tr ' ' '\n' PREFECT_HOME=/home/prefect/.prefect PREFECT_SERVER_API_HOST=192.168.29.151 PREFECT_SERVER_API_PORT=4200 PREFECT_SERVER_API_KEEPALIVE_TIMEOUT=60 PREFECT_SERVER_LOGGING_LEVEL=DEBUG PREFECT_API_DATABASE_CONNECTION_URL=postgresql+asyncpg://prefect:prefectpassword@127.0.0.1:5432/prefect PREFECT_SERVER_API_AUTH_STRING=astra:Network@5714 (.venv) root@Prefect:~/.prefect# systemctl show prefect-server -p Environment^C (.venv) root@Prefect:~/.prefect# prefect config view 🚀 you are connected to: 192.168.29.151 PREFECT_PROFILE='server' PREFECT_API_AUTH_STRING='********' (from profile) PREFECT_API_URL='192.168.29.151/api' (from profile) (.venv) root@Prefect:~/.prefect# prefect version Version: 3.7.7 API version: 0.8.4 Python version: 3.12.3 Git commit: 743957b8 Built: Thu, Jul 02, 2026 09:16 PM OS/Arch: linux/x86_64 Profile: server Server type: server Pydantic version: 2.12.5 Server: Database: sqlite SQLite version: 3.45.1 Integrations: prefect-docker: 0.7.1 (.venv) root@Prefect:~/.prefect#
m
✅ thought for 67.3 seconds
Yes — service and local CLI can absolutely be in different worlds. That is exactly what you're seeing. Why this happens 
systemctl show ... Environment
shows the environment variables for the systemd service process. But when you run:
Copy code
bash
prefect config view
prefect version
you are running Prefect in your current shell session, which only sees: - shell environment variables - Prefect profile settings - defaults It does not automatically inherit the environment from the 
prefect-server.service
. So: - systemd service environment = what the server process uses - interactive shell environment = what your CLI commands use Those are separate unless you export the same variables in your shell. That explains your output Your shell shows only:
Copy code
text
PREFECT_PROFILE='server'
PREFECT_API_AUTH_STRING=...
PREFECT_API_URL=...
So your shell-based 
prefect version
is evaluating settings from the local profile/defaults, and it falls back to SQLite. Meanwhile the running server may still be using PostgreSQL because its own systemd environment includes the DB URL. Important warning Your service environment still shows:
Copy code
text
PREFECT_API_DATABASE_CONNECTION_URL=<postgresql+asyncpg://prefect:prefectpassword@127.0.0.1:5432/prefect>
Those 
< >
are still visible there. If that is truly the literal value systemd loaded, it is suspicious. If the service is working anyway, double-check the actual unit file:
Copy code
bash
cat /etc/systemd/system/prefect-server.service
If you still see:
Copy code
ini
Environment="PREFECT_API_DATABASE_CONNECTION_URL=<postgresql+asyncpg://...>"
remove the angle brackets and restart the service. It should be:
Copy code
ini
Environment="PREFECT_API_DATABASE_CONNECTION_URL=<postgresql+asyncpg://prefect:prefectpassword@127.0.0.1:5432/prefect>"
Then:
Copy code
bash
systemctl daemon-reload
systemctl restart prefect-server
How to make your CLI see the same DB config In the same shell, export the variable manually:
Copy code
bash
export PREFECT_API_DATABASE_CONNECTION_URL="<postgresql+asyncpg://prefect:prefectpassword@127.0.0.1:5432/prefect>"
Then run:
Copy code
bash
prefect config view --show-secrets
prefect version
Now your CLI should report PostgreSQL too. Best way to verify what the running server is actually using Since your CLI and service differ, the most reliable checks are: Option 1: inspect server logs
Copy code
bash
journalctl -u prefect-server -n 100 --no-pager
Option 2: verify tables exist in PostgreSQL
Copy code
bash
psql "<postgresql://prefect:prefectpassword@127.0.0.1:5432/prefect>" -c "\dt"
If Prefect is using PostgreSQL, you should see many Prefect tables there. Option 3: check whether data is actively changing in PostgreSQL
Copy code
bash
psql "<postgresql://prefect:prefectpassword@127.0.0.1:5432/prefect>" -c "SELECT COUNT(*) FROM flow_run;"
If you create new runs in the UI and that count increases, the server is using PostgreSQL. Option 4: test whether SQLite file is no longer active Check the old SQLite DB file timestamp/size:
Copy code
bash
ls -lh /home/prefect/.prefect/prefect.db
If PostgreSQL is now the live DB, this file should stop changing. Why 
prefect version
is misleading here 
prefect version
is not asking the running server "what DB are you using?" It is evaluating configuration in the current CLI process. So in your situation: - 
prefect version
from shell → SQLite - running server process via systemd → maybe PostgreSQL That is normal when configs differ. If you want CLI and service to match You have a few options: 1. Export the env var in your shell each time
Copy code
bash
export PREFECT_API_DATABASE_CONNECTION_URL="<postgresql+asyncpg://prefect:prefectpassword@127.0.0.1:5432/prefect>"
2. Put it in your shell profile For example in `~/.bashrc`:
Copy code
bash
export PREFECT_API_DATABASE_CONNECTION_URL="<postgresql+asyncpg://prefect:prefectpassword@127.0.0.1:5432/prefect>"
3. Store it in the active Prefect profile
I’d be cautious with secrets there, but it can be done if you want the CLI always aligned. Bottom line Yes: - your systemd service and your interactive CLI are currently using different configuration sources - that is why 
prefect config view
and 
prefect version
do not match the running service - to confirm PostgreSQL is truly live, check PostgreSQL tables directly, not just local CLI output Please run these 3 commands and paste the output if you want me to confirm 100%:
Copy code
bash
cat /etc/systemd/system/prefect-server.service
psql "<postgresql://prefect:prefectpassword@127.0.0.1:5432/prefect>" -c "\dt"
ls -lh /home/prefect/.prefect/prefect.db
s
@Marvin why docker worker connected to prefect server without auth env root@Prefect:/home/prefect/worker# cat /etc/systemd/system/prefect-worker.service [Unit] Description=Prefect Server After=network.target [Service] User=root WorkingDirectory=/ ExecStart=/home/prefect/.venv/bin/prefect worker start --pool "docker-worker" Restart=always [Install] WantedBy=multi-user.target root@Prefect:/home/prefect/worker# cat /etc/systemd/system/prefect-server.service [Unit] Description=Prefect Server After=network-online.target postgresql.service Wants=network-online.target [Service] Type=simple User=root WorkingDirectory=/home/prefect Environment="PREFECT_HOME=/home/prefect/.prefect" Environment="PREFECT_SERVER_API_HOST=192.168.29.151" Environment="PREFECT_SERVER_API_PORT=4200" Environment="PREFECT_SERVER_API_KEEPALIVE_TIMEOUT=60" Environment="PREFECT_SERVER_LOGGING_LEVEL=DEBUG" Environment="PREFECT_API_DATABASE_CONNECTION_URL=postgresql+asyncpg://prefect:prefectpassword@127.0.0.1:5432/prefect" Environment="PREFECT_SERVER_API_AUTH_STRING=astra:Network@5714" ExecStart=/home/prefect/.venv/bin/prefect server start --host 0.0.0.0 --port 4200 --keep-alive-timeout 60 Restart=on-failure RestartSec=10 StandardOutput=journal StandardError=journal [Install] WantedBy=multi-user.target root@Prefect:/home/prefect/worker#
m
✅ thought for 149.0 seconds
Because the worker is probably getting its connection settings from your Prefect profile or user config, not from the worker service file. Your worker service has no env vars:
Copy code
ini
ExecStart=/home/prefect/.venv/bin/prefect worker start --pool "docker-worker"
But 
prefect worker start
still reads config from the normal Prefect settings sources, including: - environment variables - 
.env
- Prefect config/profile files So if this machine already has a profile like: - 
PREFECT_API_URL=<http://192.168.29.151:4200/api>
- 
PREFECT_API_AUTH_STRING=astra:Network@5714
then the worker can connect even though they are not written in the systemd unit. Most likely explanation in your case Earlier your CLI showed:
Copy code
text
PREFECT_PROFILE='server'
PREFECT_API_AUTH_STRING='****' (from profile)
PREFECT_API_URL='<http://192.168.29.151:4200/api>' (from profile)
That strongly suggests the worker is inheriting the same local Prefect profile/config as the 
root
user. So: - server service uses its own systemd env - worker process likely uses 
~/.prefect/profiles.toml
or root user config - that's why it works without explicit 
Environment=
lines Important distinction These two are different: - 
PREFECT_SERVER_API_AUTH_STRING
- used by the server to require Basic Auth on incoming requests - 
PREFECT_API_AUTH_STRING
- used by clients/workers to send Basic Auth to the server So if the server has:
Copy code
ini
Environment="PREFECT_SERVER_API_AUTH_STRING=astra:Network@5714"
then the worker must somehow have:
Copy code
ini
PREFECT_API_AUTH_STRING=astra:Network@5714
It does not need to be in the worker service file if it already exists in the active Prefect profile. How to confirm exactly where the worker is getting it from Run these as the same user that runs the worker (
root
in your case):
Copy code
bash
prefect config view --show-secrets
and:
Copy code
bash
cat /root/.prefect/profiles.toml
Also check whether shell env vars exist:
Copy code
bash
env | grep PREFECT
How to see the worker service environment specifically
Copy code
bash
systemctl show -p Environment --value prefect-worker | tr ' ' '\n'
If that comes back empty or without auth/url, then the worker is not getting config from systemd; it is getting it from Prefect profile/config files. Recommended setup Even if it works now, I recommend making the worker service explicit so it is predictable. Example:
Copy code
ini
[Unit]
Description=Prefect Worker
After=network.target

[Service]
User=root
WorkingDirectory=/home/prefect
Environment="PREFECT_API_URL=<http://192.168.29.151:4200/api>"
Environment="PREFECT_API_AUTH_STRING=astra:Network@5714"
ExecStart=/home/prefect/.venv/bin/prefect worker start --pool docker-worker
Restart=always

[Install]
WantedBy=multi-user.target
Then reload/restart:
Copy code
bash
systemctl daemon-reload
systemctl restart prefect-worker
One more important thing Your server service still appears to contain:
Copy code
ini
Environment="PREFECT_API_DATABASE_CONNECTION_URL=<postgresql+asyncpg://...>"
Please remove the 
< >
there if they are truly in the file. Useful docs - Prefect settings: Prefect settings - Prefect workers: Prefect workers - Prefect server: Prefect Server If you want, I can help you inspect: 1. the active 
profiles.toml
, and 2. a cleaned-up 
prefect-worker.service
and 
prefect-server.service
pair.
Open in Slack
PreviousNext
Powered by