Skip to content

ERROR: could not access file "$libdir/timescaledb-2.22.0": No such file or directory #601

@walrusmustgo

Description

@walrusmustgo

Hello! i'm encountering really unpredictable issues with these docker images. So, i've been using timescaledb/timescaledb-ha:pg17-all for a good while now ( almost a year ). Then, a few weeks ago, i've redeployed my stack, which pulled this image,and i've got the error:

ERROR: could not access file "$libdir/timescaledb-2.22.0": No such file or directory

So, i've went to docker-hub, and found the pg17-ts2.22-all image << changed the image to this one, and it worked for a few weeks. Now, i've redeployed my stack again, and the same image threw the same error again. I've then tried upgrading to pg17.6-ts2.22.1-all, and got the same error, then tried downgrading to pg17-all again, got the same error, and finally pulled pg17.6-ts2.22.0-all ,and got it working again.

Now, i'm really unsure what could be the problem here, but it's a tricky one to have :)

Y'all got any ideas perhaps?

my docker-compose entry is:

  timescaledb:
    image: timescale/timescaledb-ha:pg17.6-ts2.22.0-all
    container_name: timescaledb
    environment:
      - POSTGRES_DB=${DB_NAME}
      - POSTGRES_USER=${DB_USER}
      - POSTGRES_PASSWORD=${DB_PASS}
      - TIMESCALEDB_DATABASE=${DB_NAME}
      - TIMESCALEDB_USER=${DB_USER}
      - TIMESCALEDB_PASSWORD=${DB_PASS}
      - TIMESCALEDB_TELEMETRY=off
      - TS_TUNE_MAX_CONNS=10000
      - PGDATA=/pgdata
    depends_on:
      - timescaledb-init
    networks:
      - index_db_net
    volumes:
      - timescaledb_data:/pgdata
    command: postgres -p ${DB_PORT}
    restart: always
    ports:
      - "${DB_PORT}:${DB_PORT}"
    healthcheck:
      test: ["CMD-SHELL", "pg_isready -h localhost -p ${DB_PORT} -U ${DB_USER} -d ${DB_NAME}"]
      interval: 10s
      timeout: 5s
      retries: 5
      start_period: 30s
    logging:
      driver: "json-file"
      options:
        max-size: "100m"
        max-file: "3"

It's pretty basic, haven't changed this for a while either.

Cheers!

PS. the timescaledb-init container that's referenced here is a basic permission fix, nothing that could cause this.

==== EDIT ====

I reckon that i should have run

ALTER EXTENSION timescaledb UPDATE;

when i pinned to the 2.22.1 image? It's kinda in the docs :D

One question though, if i run this command on a mismatched image version, will it break? e.g. if i set up my init container to update the extension on every redeploy, and pin to pg17-all image, can i expect to have the extension and the image at the same version?

You can close this issue if you think everything is in order. :)

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type
    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions