Обсуждение: [GENERAL] Invalid magic number 0000 in log file
Hi everyone, a customer reported an issue starting Postgres on his machine after reinstalling Windows (10 before, 10 now) without formatting (so there's Windows and Windows.old) because of Windows not starting up anymore. Before this reinstallation everything was OK. Now here's the log 2017-08-25 19:34:49 CEST LOG: database system was shut down at 2017-08-25 08:37:08 CEST 2017-08-25 19:34:49 CEST LOG: invalid magic number 0000 in log file 1, segment 53, offset 0 2017-08-25 19:34:49 CEST LOG: invalid primary checkpoint record 2017-08-25 19:34:49 CEST LOG: invalid secondary checkpoint record 2017-08-25 19:34:49 CEST PANIC: could not locate a valid checkpoint record This application has requested the Runtime to terminate it in an unusual way. Please contact the application's support team for more information. 2017-08-25 19:34:50 CEST LOG: startup process (PID 6652) exited with exit code 3 2017-08-25 19:34:50 CEST LOG: aborting startup due to startup process failure PostgreSQL is 9.1.6 32-bit. (I know, EOLed, but it's hard to keep up to date about 1000 customer machines!) I know that's related to XLOG files, but do not have idea on how to deal with it. I had a quick googling but found nothing but "your files are corrupted. You'd better initdb." Then I bumped into this (old) post from Alvaro https://www.postgresql.org/message-id/20070515173401.GB12731@alvh.no-ip.org If it's applicable, is it still valid or too many things have changed? Thanks in advance, Moreno.
On 8/25/17 14:24, Moreno Andreo wrote: > I know that's related to XLOG files, but do not have idea on how to deal > with it. I had a quick googling but found nothing but "your files are > corrupted. You'd better initdb." That's pretty much it. You might not have to initdb if you have an uncorrupted copy of the WAL files somewhere (i.e., a backup). But then I would have concerns about what else might be corrupted. > Then I bumped into this (old) post from Alvaro > > https://www.postgresql.org/message-id/20070515173401.GB12731@alvh.no-ip.org > > If it's applicable, is it still valid or too many things have changed? That doesn't seem related. -- Peter Eisentraut http://www.2ndQuadrant.com/ PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services
Il 25/08/2017 22:01, Peter Eisentraut ha scritto: > On 8/25/17 14:24, Moreno Andreo wrote: >> I know that's related to XLOG files, but do not have idea on how to deal >> with it. I had a quick googling but found nothing but "your files are >> corrupted. You'd better initdb." > That's pretty much it. Thanks, I'll report to the customer > > You might not have to initdb if you have an uncorrupted copy of the WAL > files somewhere (i.e., a backup). But then I would have concerns about > what else might be corrupted. Probably system was not booting because of a faulty hard disk, and the fault also involved some WAL files. Once reinstalled, the system marked as unusable the faulty sectors and everything is fine now, except for what was corrupted before.......... I'll tell him to have a deep check of OS events and of the disk itself. Thanks Moreno.-