Re: Archive recovery won't be completed on some situation. - Mailing list pgsql-hackers

From Kyotaro HORIGUCHI
Subject Re: Archive recovery won't be completed on some situation.
Date
Msg-id CAM103Dv1tbj=L6UvCS+892b_EJW=gwUsJayeTnXmSWxSKJ3s2Q@mail.gmail.com
Whole thread Raw
In response to Re: Archive recovery won't be completed on some situation.  (Heikki Linnakangas <hlinnakangas@vmware.com>)
Responses Re: Archive recovery won't be completed on some situation.  (Kyotaro HORIGUCHI <kyota.horiguchi@gmail.com>)
Re: Archive recovery won't be completed on some situation.  (Heikki Linnakangas <hlinnakangas@vmware.com>)
List pgsql-hackers
<p dir="ltr">Thank you.<p dir="ltr">2014/03/14 19:42 "Heikki Linnakangas" <<a
href="mailto:hlinnakangas@vmware.com">hlinnakangas@vmware.com</a>>:<br/> ><br /> > On 03/14/2014 12:32 PM,
KyotaroHORIGUCHI wrote:<br /> >><br /> >> Restarting server with archive recovery fails as following
just<br/> >> after it was killed with SIGKILL after pg_start_backup and some<br /> >> wal writes but before
pg_stop_backup.<br/> >><br /> >> | FATAL:  WAL ends before end of online backup<br /> >> | HINT:
Onlinebackup started with pg_start_backup() must be<br /> >> |  ended with pg_stop_backup(), and all WAL up to
thatpoint must<br /> >> |  be available at recovery.<br /> >><br /> >> What the mess is once entering
thissituation, I could find no<br /> >> formal operation to exit from it.<br /> >><p dir="ltr">> If you
killthe server while a backup is in progress, the backup is broken. It's correct that the server refuses to start up
fromthe broken backup. So basically, don't do that.<p dir="ltr">Hmm.. What I did is simplly restarting server  just
aftera crash but the server was accidentially in backup mode. No backup copy is used. Basically, the server is in the
samesituation with the simple restart after crash. The difference here is the restarting made the database completly
uselesswhile it had been not. I wish to save the database for the case and I suppose it so acceptable.<p
dir="ltr">regards,<br/> -- <br /> Kyotaro Horiguchi<br /> NTT Opensource Software Center<br /> 

pgsql-hackers by date:

Previous
From: Pavel Stehule
Date:
Subject: Re: plpgsql.warn_shadow
Next
From: Kyotaro HORIGUCHI
Date:
Subject: Re: Archive recovery won't be completed on some situation.