Re: Adjustment of spinlock sleep delays - Mailing list pgsql-hackers

From Tom Lane
Subject Re: Adjustment of spinlock sleep delays
Date
Msg-id 18059.1060179096@sss.pgh.pa.us
Whole thread Raw
In response to Re: Adjustment of spinlock sleep delays  ("Mendola Gaetano" <mendola@bigfoot.com>)
List pgsql-hackers
"Mendola Gaetano" <mendola@bigfoot.com> writes:
> What about use the same algorithm used in ethernet when a collision is
> detected?

Random backoff is what Rod suggested, but I don't care for the ethernet
method in detail, because it allows for only a fairly small number of
retries before giving up.  It's okay for an ethernet to drop packets
under load ... it's not okay for us to fail to acquire a spinlock.
        regards, tom lane


pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: DROP TYPE/DROP DOMAIN
Next
From: Tom Lane
Date:
Subject: Re: test beta build