Re: Review: Hot standby - Mailing list pgsql-hackers

From Pavan Deolasee
Subject Re: Review: Hot standby
Date
Msg-id 2e78013d0811210047t1309ee1fr44e726ad793f50ad@mail.gmail.com
Whole thread Raw
In response to Review: Hot standby  ("Pavan Deolasee" <pavan.deolasee@gmail.com>)
Responses Re: Review: Hot standby  ("Pavan Deolasee" <pavan.deolasee@gmail.com>)
Re: Review: Hot standby  (Simon Riggs <simon@2ndQuadrant.com>)
List pgsql-hackers
<br />I wonder if there is corner case below where there are no WAL records to replay during standby recovery.
Specifically,that may cause IsRunningXactDataValid() to return false since latestObservedXid remains set to
InvalidTransactionIdand that prevents postmaster from serving read-only clients.<br /><br />I don't have a test case,
butI recall seeing the issue while experimenting. Though that could be because of the WALInsertLock issue reported
earlier.<br/><br />           /*<br />                 * Can we signal Postmaster to enter consistent recovery mode?<br
/>                 *<br />                 * There are two points in the log that we must pass. The first<br
/>                * is minRecoveryPoint, which is the LSN at the time the<br />                 * base backup was taken
thatwe are about to rollfoward from.<br />                  * If recovery has ever crashed or was stopped there is
also<br/>                 * another point also: minSafeStartPoint, which we know the<br />                 * latest LSN
thatrecovery could have reached prior to crash.<br />                  *<br />                 * We must also have
assembledsufficient information about<br />                 * transaction state to allow valid snapshots to be
taken.<br/>                 */<br />                if (!reachedSafeStartPoint &&<br />                     
IsRunningXactDataValid()&&<br />                     XLByteLE(ControlFile->minSafeStartPoint, EndRecPtr)
&&<br/>                     XLByteLE(ControlFile->minRecoveryPoint, EndRecPtr))<br />                 {<br
/>                   reachedSafeStartPoint = true;<br />                    if (InArchiveRecovery)<br
/>                   {<br />                        ereport(LOG,<br />                            (errmsg("database has
nowreached consistent state at %X/%X",<br />                                 EndRecPtr.xlogid, EndRecPtr.xrecoff)));<br
/>                       InitRecoveryTransactionEnvironment();<br />                       
StartCleanupDelayStats();<br/>                        if (IsUnderPostmaster)<br />                            
SendPostmasterSignal(PMSIGNAL_RECOVERY_START);<br/>                    }<br />                }<br /><br /><br
/>Thanks,<br/>Pavan<br clear="all" /><br />-- <br />Pavan Deolasee<br />EnterpriseDB     <a
href="http://www.enterprisedb.com">http://www.enterprisedb.com</a><br/> 

pgsql-hackers by date:

Previous
From: "Pavan Deolasee"
Date:
Subject: Review: Hot standby
Next
From: Magnus Hagander
Date:
Subject: Re: Should enum GUCs be listed as such in config.sgml?