Dbeaver: Canceling the query takes forever

[antonyr]

My DB is on VPN. I accidentally hit execute query, later I came to know I didn't connect to VPN. Now I decide to cancel the query in execution. I cancelled it, but still it can't cancel it. It takes forever now, before it can stop execution.

[allanlaal]

same here on 5.0.2 with VPN - 942s and counting..

only killing DBeaver helps :(

[MarcinWieczorek]

Confirmed. Is there a solution?

[alex-ant]

Same thing when working with local databases.

[serge-rider]

IT is a bug in 5.2.2, will be fixed in 5.2.3 (fix is already in Early Access - https://dbeaver.io/files/ea)

[cloud-66]

Hello, i use dbeaver for oracle database, but cancellation of query don't working. I see thig bug in version 5.2.5 . it was not fixed in 5.2.3, i don't see any imrovements.
I like this apps but some problems are very frustrating.

[reichg]

Confirmed cloud-66s observation that this bug still exists in 5.2.5. If you try to cancel a long running query it will not cancel unless you close dbeaver.

[MFernstrom]

Still happening in 5.3.4 (Feb 3, 2019 release)

[jordanmat]

Still happening in 6.0.0.201903101945

[drmaquino]

still happening in version 6.0.3 ......

[Dania02525]

Still a problem in 5.3.3:

[suricat89]

Same problem here with Oracle Database on Dbeaver 6.1.1.201906240635

[larvinloy]

Same issue with Redshift queries. Redshift console shows that the query has already been cancelled, but DBeaver is still stuck trying to cancel the query. Occasionally it causes the app to crash.

v 6.0.5

[namansachdeva]

Still happening in 6.1.4

[serge-rider]

See #3616
Generally query cancel support is provided by database driver.
Some drivers just don't support it. We can't fix it from DBeaver side :(

[namansachdeva]

Force quiting DBeaver and restarting it solves the problem. It reconnects
the database. Maybe a hack to resolve the problem?

On Tue, 17 Sep 2019, 3:42 am Serge Rider, [e-mail] wrote:

See #3616 https://github.com/dbeaver/dbeaver/issues/3616
Generally query cancel support is provided by database driver.
Some drivers just don't support it. We can't fix it from DBeaver side :(

—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
https://github.com/dbeaver/dbeaver/issues/3094?email_source=notifications&email_token=AEMHLD6NEJGDRTFSCP6NHADQKAADBA5CNFSM4ET33KO2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD62VQPI#issuecomment-531978301,
or mute the thread
https://github.com/notifications/unsubscribe-auth/AEMHLD43BHCK6WOTCWRUADTQKAADBANCNFSM4ET33KOQ
.

[rdp]

See also https://github.com/dbeaver/dbeaver/issues/6755

[barretod]

This is still broken on 6.2.3, very annoying.

[DanielWFrancis]

still broken interfacing with postgres 11 on AWS RDS

[anveshragi]

This is still happening on 6.2.5 with AWS Redshift as source

[mikkk3l]

Still happens in 6.3.2 with ojdbc7.jar

[jedeleh]

Same. Honestly I've always had this issue but it wasn't until recently that the "cancel query" spun forever until I restart dbeaver

[janpxr]

Same issue with:
dbeaver-ce-6.3.5-x86_64
and
dbeaver-ce-7.0.0-x86_64

Even in the cases where query never gets to the database or has finished really quick there and for some reason Dbeaver does not get the error/resultset back before you click the "Cancel" button then it will keep showing that it is executing. Example:

So it is not related to the database drivers and issues in those cases, just the Cancel button hangs and there is nothing else you can do with it than restart Dbeaver.

This kind of issue over several versions and in 2020, really?

[TimonKK]

Same issue with dbeaver-ce-7.0.1-x86_64 on Ubunti 19.10

[archismandinda]

Issue still persists in Version 7.0.0.202003021717.

The greater issue here is that if there is a valid long running query which you want to cancel, even if you close DBeaver and restart the app, the query still keeps getting executed on the database.
Restarting DBeaver doesn't actually kill the query.

So many times I get flagged by my DB Admins because some long running query would be running even though I would have thought that they got cancelled since I restarted DBeaver.

[disk-edit]

Issue still in latest stable build.
Running 7.0.3.202004191823 on Ubuntu 18.04

[akbarumar88]

Issue still persis in 7.0.3 it was really annoyying, do you guys have another recommendations for multi DBMS Database tool?

[janpxr]

@archismandinda the initial issue here is not about getting the database session killed (as in the described case the VPN was not connected then the query never got to the database anyway; to get you database sessions killed then there are separate issues for that; also if in your case the sessions stay active on the database side then it might not be DBeavers fault - it might be also be in the JDBC driver, network issues etc.).

So lets just keep the issue here concentrated to the issue that if I hit the cancel button then the expected result is at least that I would not need to restart the DBeaver application.

@akbarumar88 there are many such tools around SQuirreL SQL Client, DbVisualizer, DataGrip etc. All of them have issues alongside with great functionality. Still DBeaver is i general great tool, just then with its own issues.

[joshcrichman]

Can we please un-close this issue? It's clearly not resolved in any way. Definitely still an issue for me. Always has been and feels like always will be...

[serge-rider]

Well, it is tricky.

Cancel works for PostgreSQL, MySQL, some Oracles and for some other databases.
It doesn't work for, e.g., Redshift, some Oracle, etc.

Cancel must be supported by the database server and by the driver. We can't perform cancel on client side.
For drivers which don't support cancel we wait several seconds and then try to disconnect. It works for some drivers as disconnect cancels any active queries.
But some other drivers (e.g. Redshift) don't allow to disconnect when a query is running. It just hangs in a mutex.
For some databases even restarting client application doesn't help. They just do not support query interruption at all. Once started query will run until client session will be terminated by admin or server will be restarted.

If you have any ideas or any workarounds on mind - please share or create a new ticket.

Probably this should be resolved for particular databases individually in separate tickets.

[mpaloni]

Facing the issue with Redshift.

Would it be possible to crash only the connection if the query takes too long(or connection is on redshift database) and then restoring it? For example RStudio & jupyter notebook has the kernel running as seperate and that can be restarted as many times as you want without crashing the application itself.

(This suggestion comes with 1% of knowledge about dbeaver)

[andrewkp101]

Dbeaver Version 7.1.1.202006211844 MySQL has this issue still.

Just for example MySQL Workbench with the same query and same database cancels correctly.

By doing
show full processlist

I can still see this process "Sending data"

Given that MySQL Workbench works fine in cancelling an active mysql query suggests Dbeaver might be improved.

[carlosfg97]

As many people mention, this is obviously an ongoing issue that bothers a lot. Are we going to receive any news from the developers?

[zsaltys]

This is a very annoying issue. Would be great if devs would acknowledge it. Having a way to force kill all background tasks would be good enough.

[uslss]

Issue has been fixed in scope of 7.1.3

9121

[joshcrichman]

It said it was fixed in the 7.1.2 release notes. Definitely was not for me
using Redshift. I was supremely disappointed.
[image: image.png]

On Wed, Jul 22, 2020 at 10:04 AM uslss [e-mail] wrote:

Issue has been fixed in scope of 7.1.3

9121 https://github.com/dbeaver/dbeaver/issues/9121

—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
https://github.com/dbeaver/dbeaver/issues/3094#issuecomment-662506714,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/AGKDMCFWVRA3T3YVNIR6ZBLR435YDANCNFSM4ET33KOQ
.

[joshcrichman]

See above - still having issues with this using Redshift (and maybe others, since other people are still complaining). Can we open this ticket again, please?

[uslss]

@joshcrichman , could you create a new ticket with more detailed description of the problem you're observing?
Thanks in advance :)

[joshcrichman]

Sure, though it's exactly the same as the many other tickets already filed,
and I don't want to lose all the people that have commented here and
elsewhere to show how relevant it is.

On Wed, Jul 22, 2020 at 3:06 PM uslss [e-mail] wrote:

@joshcrichman https://github.com/joshcrichman , could you create a new
ticket with more detailed description of the problem you're observing?
Thanks in advance :)

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
https://github.com/dbeaver/dbeaver/issues/3094#issuecomment-662668582,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/AGKDMCB7ICWDDOJZRBVZAI3R45BCRANCNFSM4ET33KOQ
.

[jpierre]

I also have the same issue with Redshift

[yuhadooper]

I've been facing the same issue with redshift/dbeaver from the past 3 years!

[nguyenvinhlinh]

August 4, 2020, I got this issue. Please help.

[pfeiffee]

Does anyone know if there are any workarounds like driver properties and timeouts that could be set to at least be able to get away with waiting a minute instead of force killing dbeaver?

[christopher-bonitz]

Used to have the same issue (since I always connect to client systems through VPN) but in recent 7.2 the problem is gone.
If someone could re-test, to see if this issue is still valid, it would be helpful I think.

[andrewkp101]

Previously I had more luck with native as opposed to java drivers, however I did not test this heavily and am not using dbeaver at the current time. I think anyone facing this problem should try a stable native driver and see if it improves the situation.

Get Outlook for Androidhttps://aka.ms/ghei36

---

From: Christopher Bonitz [e-mail]
Sent: Monday, September 7, 2020 2:11:51 PM
To: dbeaver/dbeaver [e-mail]
Cc: andrewkp101 [e-mail]; Comment [e-mail]
Subject: Re: [dbeaver/dbeaver] Canceling the query takes forever (#3094)

Used to have the same issue (since I always connect to client systems through VPN) but in recent 7.2 the problem is gone.
If someone could re-test, to see if this issue is still valid, it would be helpful I think.

—
You are receiving this because you commented.
Reply to this email directly, view it on GitHubhttps://github.com/dbeaver/dbeaver/issues/3094#issuecomment-688316565, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ANQPD575MBG67GLNB4SDMGTSETLZPANCNFSM4ET33KOQ.

[disk-edit]

I have had this same issue for many versions with queries run against AWS RDS MariaDB and AWS Redshift.

Now I am running DBeaver 7.2.0 with RedshiftJDBC42-no-awssdk-1.2.45.1069.jar in play. I was able to immediately cancel queries run against AWS Redshift. I also had success canceling queries against AWS RDS MariaDB.

Great progress! Thank you.

[joshcrichman]

I just upgraded to 7.2. Will try it out for a week or so and report back.

On Tue, Sep 8, 2020 at 8:26 AM Ben Baker [e-mail] wrote:

I have had this same issue for many versions with queries run against AWS
RDS MariaDB and AWS Redshift.

Now I am running DBeaver 7.2.0 with
RedshiftJDBC42-no-awssdk-1.2.45.1069.jar in play. I was able to immediately
cancel queries run against AWS Redshift. I also had success canceling
queries against AWS RDS MariaDB.

Great progress! Thank you.

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
https://github.com/dbeaver/dbeaver/issues/3094#issuecomment-688864660,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/AGKDMCEH2NBE3FIZXYJEFMLSEYWJDANCNFSM4ET33KOQ
.

[uslss]

fixed in 7.2

[yuhadooper]

Yep, finally fixed in the 7.2

[mikkk3l]

Thank you very much! It is fixed in 7.2 with a workaround which is to close the connection if cancel request doesn't respond within a timeout period.

There is a little bug still. If you close the connection then the blue arrows show the query as it were still executing:

This is not a big deal, just a graphic bug.

But if, at that point, you try to export a query to CSV it fails saying that you can't excecute simultaneous queries and you are not doing it. The only workaround is execute another query and, when it finishes OK, you can then export a query result to a file.

Could you please fix this issue? Thanks!