Re: Stalls on PGSemaphoreLock

От: Gudmundsson Martin (mg)
Тема: Re: Stalls on PGSemaphoreLock
Дата: ,
Msg-id: 2E5766F28426E547AF0A2A870FAE630D22EAC0@SEGOTNC5182-N2.vcn.ds.volvo.net
(см: обсуждение, исходный текст)
Ответ на: Re: Stalls on PGSemaphoreLock  (Pavy Philippe)
Список: pgsql-performance

Скрыть дерево обсуждения

Stalls on PGSemaphoreLock  (Matthew Spilich, )
 Re: Stalls on PGSemaphoreLock  (Ray Stell, )
  RE : Stalls on PGSemaphoreLock  (Pavy Philippe, )
 Re: Stalls on PGSemaphoreLock  (Matthew Spilich, )
  Re: Stalls on PGSemaphoreLock  (Pavy Philippe, )
   Re: Stalls on PGSemaphoreLock  ("Gudmundsson Martin (mg)", )
   Re: Stalls on PGSemaphoreLock  (Matthew Spilich, )
    Re: Stalls on PGSemaphoreLock  (Matheus de Oliveira, )
 Re: Stalls on PGSemaphoreLock  (Emre Hasegeli, )
 Re: Stalls on PGSemaphoreLock  (Takashi Horikawa, )

Hi Pavy!
What kernel version/RHEL release are you running on the servers you are experiencing these issues?
I'm interested in knowing since I suspect similar issues on some of our database servers.

Best regards, Martin

> -----Original Message-----
> From:  [mailto:pgsql-performance-
> ] On Behalf Of Pavy Philippe
> Sent: den 25 mars 2014 9:10
> To: Matthew Spilich; 
> Subject: Re: [PERFORM] Stalls on PGSemaphoreLock
>
> Here, we were the transparent hugepage always actif:
>         cat /sys/kernel/mm/redhat_transparent_hugepage/enabled
>         [always] never
>
> We changed to:
> cat /sys/kernel/mm/redhat_transparent_hugepage/enabled
>         always [never]
>
>
>
> For the semaphore, our initial configuration was:
>         cat /proc/sys/kernel/sem
>         250 32000 32 128
>
> And we changed to:
>         cat /proc/sys/kernel/sem
>         5010    641280  5010    128
>




В списке pgsql-performance по дате сообщения:

От: "Gudmundsson Martin (mg)"
Дата:
Сообщение: Re: Stalls on PGSemaphoreLock
От: Dave Johansen
Дата:
Сообщение: Partitions and prepared statements?