> On Mar 31, 2025, at 12:36, Eden Aharoni wrote:
> Hope you’ll have any clue 😊
Based on that, I'd take it up with AWS. It does seem that the EBS mount is
under-performing.
On 3/31/25 11:23 AM, Christophe Pettus wrote:
On Mar 31, 2025, at 10:54, Eden Aharoni wrote:
So you believe it's strictly an EBS issue?
You are asking about disk I/O, that is the purview of the OS and
hardware. In your case both are created and managed by AWS, only they
know what is g
> On Mar 31, 2025, at 11:30, Eden Aharoni wrote:
>
> Can you please tell me what other data might help?
You can show your work on how you got the megabytes/second number. (Be aware
that on a general open-source mailing list, there's only so much debugging that
we can do of a specific probl
ubject: Re: [EXTERNAL] RDS IO Read time
> On Mar 31, 2025, at 10:54, Eden Aharoni wrote:
>
> So you believe it's strictly an EBS issue?
Well, PostgreSQL certainly can read faster than 34MB/s off of disk. With the
data you've given, I can't really say if it's purely an EBS issue.
sec.. (ive done 4314
* 8kb / 1.2sec which is about 28 MB/s)…
Hope you’ll have any clue 😊
From: Christophe Pettus
Date: Monday, 31 March 2025 at 21:44
To: Eden Aharoni
Cc: pgsql-general@lists.postgresql.org
Subject: Re: [EXTERNAL] RDS IO Read time
> On Mar 31, 2025, at 11:30, Eden Aharoni
So you believe it's strictly an EBS issue?
Get Outlook for Android<https://aka.ms/AAb9ysg>
From: Christophe Pettus
Sent: Monday, March 31, 2025 8:43:19 PM
To: Eden Aharoni
Cc: pgsql-general@lists.postgresql.org
Subject: Re: [EXTERNAL] RDS IO Read
> On Mar 31, 2025, at 10:54, Eden Aharoni wrote:
>
> So you believe it's strictly an EBS issue?
Well, PostgreSQL certainly can read faster than 34MB/s off of disk. With the
data you've given, I can't really say if it's purely an EBS issue.
> On Mar 31, 2025, at 10:32, Eden Aharoni wrote:
>
> First, thanks for the reply :)
> So, I do know which part is taking a lot of IO time and it's to be honest any
> node that reads from the disk.. of course, we're running EXPLAIN on our
> queries (to be more specific we use auto_explain) bu
