<p>Thank you for the patch.<p>The test script finishes in success with that. And looks reasonable on a short
glance.<p>OnFri, Apr 26, 2013 at 4:34 AM, Heikki Linnakangas <<a
href="mailto:hlinnakangas@vmware.com">hlinnakangas@vmware.com</a>>wrote:<br /> > One idea to fix this is to not
setcurFileTLI, until the page header on the<br /> > just-opened file has been verified. Another idea is to change
thecheck in<br /> > XLogFileReadAnyTLI() that currently forbids curFileTLI from moving<br /> > backwards. We
couldallow curFileTLI to move backwards, as long as the<br /> > tli is >= ThisTimeLineID (ThisTimeLineID is the
currenttimeline we're<br /> > recovering records from).<br /> ><br /> > Attached is a patch for the 2nd
approach.With the patch, the test script<br /> > works for me. Thoughts?<p>I am uncertain a bit weather it is
necessaryto move curFileTLI to anywhere randomly read . On a short glance, the random access occurs also for reading
checkpoint-relatedrecords.<br /> Also I don't have clear distinction between lastSegmentTLI and curFileTLI after the
patchapplied. Although , I need look closer around them to understand.<p>> PS. This wasn't caused by the 9.2.4
changeto do crash recovery before<br /> > entering archive recovery. The test script fails in the same way with
9.2.3<br/> > as well.<p>--<br /> Kyotaro Horiguchi