Thread: [doc] minor fix for CREATE INDEX CONCURRENTLY

Hi,

I believe that the description of which transactions will block a CREATE
INDEX CONCURRENTLY command from proceeding isn't right.

Please find the attached patch.

Best regards,
Frédéric

Hello!

No, I think the comment is correct, it is about [0].

[0]:
https://github.com/postgres/postgres/blob/f5a987c0e5f6bbf0cc0420228dc57e7aae4d7e8f/src/backend/commands/indexcmds.c#L4217

On 7/9/25 16:06, Mihail Nikalayeu wrote:
> Hello!
> 
> No, I think the comment is correct, it is about [0].
> 
> [0]:
https://github.com/postgres/postgres/blob/f5a987c0e5f6bbf0cc0420228dc57e7aae4d7e8f/src/backend/commands/indexcmds.c#L4217

Aaahhh... yes, you're right! thanks!

On 7/9/25 16:30, Frédéric Yhuel wrote:
> 
> 
> On 7/9/25 16:06, Mihail Nikalayeu wrote:
>> Hello!
>>
>> No, I think the comment is correct, it is about [0].
>>
>> [0]: https://github.com/postgres/postgres/blob/ 
>> f5a987c0e5f6bbf0cc0420228dc57e7aae4d7e8f/src/backend/commands/ 
>> indexcmds.c#L4217
> 
> Aaahhh... yes, you're right! thanks!
> 
> 


Come to think of it, shouldn't it be that instead?

"and in addition it must wait for all existing transactions that modify 
the indexed table, and for those that could potentially modify or use 
the index to terminate."

I think the first set isn't included in the second one, as long as the 
default isolation level is READ COMMITTED.