"Detect unique identifiers" takes tens of seconds for each new query (Postgre).

When I start query execution DBeaver writes first Detect unique identifiers and spend 10-15s, then actually executes query.

There is a button “Cancel” to interrupt that process. After pushing it and around 5s delay I see alert: Database driver is not responding. Do you want to cancel request and close connection?. I chose no and got response a bit later.

What feature “Detect unique identifiers” enables? Can I disable probing of “unique identifiers”?

I cannot find such option, similar option was mentioned here: https://github.com/dbeaver/dbeaver/issues/5633

My Postgre is PostgreSQL 9.6.10 on x86_64-alpine-linux-musl, compiled by gcc (Alpine 6.3.0) 6.3.0, 64-bit running inside VirtualBox (Vagrant).

I see similar tips (other messages appear, like “Load TableCache”, etc) appear when I connect to Google Cloud Managed Postgre 11 but it takes fraction of seconds to finish.

I updated DBeaver Version 7.3.4.202101311042 from scoop install dbeaver. Same behavior is with old Version 6.2.2.201910072100

[kseniiaguzeeva]

1. Could you please see which query gives so long execution time (Window>Show view>Query manager)?

2. Try to disable Read table metadata and Read table preferences in Preferences>Data Editor

Even if disabling options helps, I will be glad to receive the query from Query manager and investigate it.

[elektro-wolle]

I’ve tried the statement directly on the database and can confirm the ten seconds. But without the pg_stat_get_numscans expression the execution is instant (50ms).