Skip to content

Help

InfluxDB, QuestDB and TimescaleDB errors and how to fix them

Tested with:Databasezy API v1 · zb CLI 0.1

TimescaleDB is available on every plan, including Free. InfluxDB 3 Core and QuestDB are available on paid plans only. All three sit behind the same gateway as every other engine. So the generic errors, such as allow-list blocks, paused instances and connection limits, look the same here as on the PostgreSQL page (for TimescaleDB and QuestDB queries) or as JSON errors with HTTP status codes (for the HTTPS APIs).

Cause: every endpoint except /health and /ping needs the token. The request has no token, the wrong token, or a token that was replaced. After a rotation, the previous token stops working immediately. There is no grace period.

Fix: send Authorization: Bearer <token>. After a rotation, wait for the rotation to report it is applied, then reveal the new token with zb instances credentials reveal <instance_id>. If you regenerate the operator token yourself with InfluxDB’s own API, the platform’s copy stops working too. Use named tokens for your applications instead.

Cause: another token operation is still pending, or a restore has not finished.

Fix: wait for the current operation to complete, then try again.

Cause: InfluxDB 3 does not support Flux.

Fix: query with SQL (/api/v3/query_sql) or InfluxQL (/api/v3/query_influxql). Flux tasks and dashboards do not carry over.

Client library queries fail, but HTTP queries work

Section titled “Client library queries fail, but HTTP queries work”

Cause: the InfluxDB 3 client libraries query over Arrow Flight, which is gRPC on the same port. A proxy on your side may block HTTP/2.

Fix: allow HTTP/2 to the host, or use the HTTP query endpoint.

Fix: use the influxdb_v2 output with the token, and the database name as the bucket. Databases are created on first write.

The InfluxDB 3 Processing Engine is not offered on Databasezy.

Cause: the user is admin with the generated password, the same for the PostgreSQL wire and the HTTP API. After a rotation, the server restarts and the old password stops working. There is no overlap.

Fix: reveal the current password and update both your query and ingestion clients.

Cause: usually the wrong port or database name.

Fix: use port 5432, database qdb and sslmode=verify-full. The HTTP API (REST, CSV import, line protocol over HTTP and the web console) is on port 443 of the same host, or 8443 in single-IP regions.

An ingestion client using tcp:: cannot connect

Section titled “An ingestion client using tcp:: cannot connect”

Cause: line protocol over TCP (port 9009) is not exposed.

Fix: use the HTTPS transport: https::addr=<host>:443;username=admin;password=...;.

Cause: QuestDB implements the PostgreSQL wire protocol, but not all of PostgreSQL’s SQL.

Fix: use drivers and BI tools for queries, and create tables with QuestDB SQL or CSV import.

TimescaleDB is PostgreSQL with the extension preinstalled. Connection and login errors are the same as on the PostgreSQL errors page.

A licence error when using compression, continuous aggregates or policies

Section titled “A licence error when using compression, continuous aggregates or policies”

Cause: Databasezy runs the Apache-2.0 edition. These Timescale License features are not available:

  • native compression and the columnstore
  • continuous aggregates
  • the job scheduler and its policies, including add_job
  • time_bucket_gapfill
  • the Toolkit hyperfunctions

Fix: use materialized views refreshed by your own scheduler, and drop old data with drop_chunks().

permission denied to set parameter "timescaledb.restoring"

Section titled “permission denied to set parameter "timescaledb.restoring"”

Cause: you ran Timescale’s own dump-and-restore recipe. It needs a superuser, and the generated app user is not one.

Fix: use zb migrate or the portal’s Migrate in tab. These run the restore steps with a short-lived superuser credential. See Migrate into TimescaleDB.

TimescaleDB restore needs a superuser connection ...

Section titled “TimescaleDB restore needs a superuser connection ...”

Cause: this is a known limitation. Restoring a backup into a new TimescaleDB instance needs superuser access that the platform does not grant yet.

Fix: restore in place instead, or use zb migrate. See Restore a backup.

create extension timescaledb fails on a PostgreSQL instance

Section titled “create extension timescaledb fails on a PostgreSQL instance”

Cause: TimescaleDB ships as its own engine, not as an extension of the PostgreSQL engine.

Fix: create a TimescaleDB instance and move the data with zb migrate --from <postgres-instance> --to <timescale-instance>.

Search the help center or ask the assistant in the portal Help panel. If that does not solve it, open a support ticket from the portal (Help → Open a ticket) and include:

  • the instance id, engine and region
  • the endpoint or port you used
  • the exact error text or HTTP status and body
  • your client, agent or driver and its version
  • the time of the failure in UTC

Never paste passwords, tokens, connection strings with passwords, API keys or patient data (PHI) into a ticket. See How to get help.