CIC and deadlocks - Mailing list pgsql-hackers

From Pavan Deolasee
Subject CIC and deadlocks
Date
Msg-id 2e78013d0703310515v443153efi49da147cc59780c7@mail.gmail.com
Whole thread Raw
Responses Re: CIC and deadlocks  ("Simon Riggs" <simon@2ndquadrant.com>)
Re: CIC and deadlocks  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers
<br />Isn't CREATE INDEX CONCURRENTLY prone deadlock conditions ?<br />I saw one with VACUUM today. But I think it can
happenwith other<br />commands like VACUUM FULL, CLUSTER, CREATE INDEX<br />CONCURRENTLY and so on. These commands
conflicton the <br />ShareUpdateExclusiveLock held by CIC and hence would wait for<br />CIC to release the lock. At the
sametime, CIC would wait for these<br />transactions to complete.<br clear="all" /><br />We know that these commands
arerun in a separate transaction <br />and do not do any index fetches or inserts/updates. So in principle<br />CIC
neednot wait for these transactions to complete in any<br />of its waits. May be we can skip waits on the transactions
that<br/>are running one of these commands. <br /><br />Is it something worth doing  ?<br /><br />Thanks,<br />Pavan<br
/><br/>-- <br /><br />EnterpriseDB     <a href="http://www.enterprisedb.com">http://www.enterprisedb.com</a> 

pgsql-hackers by date:

Previous
From: Russell Smith
Date:
Subject: Re: Feature thought: idle in transaction timeout
Next
From: "Simon Riggs"
Date:
Subject: Re: CIC and deadlocks