From: | srinivas oguri <srinivasoguri7(at)gmail(dot)com> |
---|---|
To: | Bert <biertie(at)gmail(dot)com> |
Cc: | pgsql-admin <pgsql-admin(at)postgresql(dot)org> |
Subject: | Re: PostgreSQL 12 VS PostgreSQL 15 |
Date: | 2023-07-17 18:03:32 |
Message-ID: | CADfH0yvFcMZOXkHtYZqAegoHWN+thXO1mAZnNPYM7C+6L-FMDQ@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-admin |
Yes all settings are same.
On Mon, Jul 17, 2023, 9:51 PM Bert <biertie(at)gmail(dot)com> wrote:
> Oops. Should have put my glasses on :-)
> Are all other settings the same?
>
> On Mon, 17 Jul 2023 at 18:13, srinivas oguri <srinivasoguri7(at)gmail(dot)com>
> wrote:
>
>> Thanks for responding.
>>
>> In 12:
>>
>> Number of transactions: 863125
>> TPS : 959.153959
>>
>> In 15:
>>
>> Number of transactions: 691414
>> TPS: 768.337953
>>
>> Thanks
>> Srinivas
>>
>> On Mon, Jul 17, 2023, 9:10 PM Bert <biertie(at)gmail(dot)com> wrote:
>>
>>> not an expert here.
>>> But it seems you processed almost 20K extra transactions. How is that a
>>> slowdown? :)
>>>
>>> Bert
>>>
>>> On Mon, Jul 17, 2023 at 4:19 PM srinivas oguri <srinivasoguri7(at)gmail(dot)com>
>>> wrote:
>>>
>>>> Hi Experts,
>>>>
>>>> I am doing performance testing on Aurora PostgreSQL version 12, 15
>>>> using pgbench. I see the performance is low in 15 compared to 12.
>>>>
>>>> Below is the hardware configuration:
>>>>
>>>> r6g.large
>>>> 16 GB of RAM
>>>> 2 vcpu
>>>>
>>>> Below is the command used to initialize the tables: (I have used the
>>>> pgbench of 15 version to generate the partitioned tables as we have more
>>>> partitioned tables).
>>>>
>>>> ./pgbench --initialize --scale=1000 --partition-method=range
>>>> --partitions=30 -h <>.rds.amazonaws.com -U postgres -d postgres
>>>>
>>>> For PostgreSQL 15:
>>>>
>>>> ./pgbench --client=16 --time=900
>>>> --host=<>.rds.amazonaws.com -U postgres -d postgres
>>>> pgbench (15.3)
>>>> transaction type: <builtin: TPC-B (sort of)>
>>>> scaling factor: 1000
>>>> partition method: range
>>>> partitions: 30
>>>> query mode: simple
>>>> number of clients: 16
>>>> number of threads: 1
>>>> maximum number of tries: 1
>>>> duration: 900 s
>>>> number of transactions actually processed: 691414
>>>> number of failed transactions: 0 (0.000%)
>>>> latency average = 20.824 ms
>>>> initial connection time = 151.272 ms
>>>> *tps = 768.337953 (without initial connection time)*
>>>> For PostgreSQL 12:
>>>>
>>>> ./pgbench --client=16 --time=900
>>>> --host=<>.rds.amazonaws.com -U sa -d postgres
>>>> pgbench (15.3, server 12.11)
>>>> transaction type: <builtin: TPC-B (sort of)>
>>>> scaling factor: 1000
>>>> partition method: range
>>>> partitions: 30
>>>> query mode: simple
>>>> number of clients: 16
>>>> number of threads: 1
>>>> maximum number of tries: 1
>>>> duration: 900 s
>>>> number of transactions actually processed: 863125
>>>> number of failed transactions: 0 (0.000%)
>>>> latency average = 16.681 ms
>>>> initial connection time = 146.719 ms
>>>> *tps = 959.153959 (without initial connection time)*
>>>>
>>>> I did testing multiple times with different connections but still the
>>>> same.
>>>>
>>>> Thanks
>>>> Srinivas
>>>>
>>>
>>>
>>> --
>>> Bert Desmet
>>> 0477/305361
>>>
>> --
> Bert Desmet
> 0477/305361
>
From | Date | Subject | |
---|---|---|---|
Next Message | hubert depesz lubaczewski | 2023-07-18 10:42:32 | Re: Help with logical replication: pg_replication_slots.active_pid is null???? |
Previous Message | Laurenz Albe | 2023-07-17 16:47:43 | Re: Seeking Advice on Table Alterations without Downtime in PostgreSQL |