Re: [HACKERS] "using previous checkpoint record at" maybe not the greatest idea?

On Tue, Feb 2, 2016 at 5:28 AM, Andres Freund wrote: > > Hi, > > currently if, when not in standby mode, we can't read a checkpoint > record, we automatically fall back to the previous checkpoint, and start > replay from there. > > Doing so without user intervention doesn't

Re: [HACKERS] "using previous checkpoint record at" maybe not the greatest idea?

On Sun, Feb 7, 2016 at 10:54 AM, Amit Kapila wrote: > > On Tue, Feb 2, 2016 at 5:28 AM, Andres Freund wrote: > > > > Hi, > > > > currently if, when not in standby mode, we can't read a checkpoint > > record, we automatically fall back to the previous

Re: [HACKERS] "using previous checkpoint record at" maybe not the greatest idea?

David G. Johnston wrote: > ​Learning by reading here... > > http://www.postgresql.org/docs/current/static/wal-internals.html > """ > ​After a checkpoint has been made and the log flushed, the checkpoint's > position is saved in the file pg_control. Therefore, at the start of > recovery, the

Re: [HACKERS] "using previous checkpoint record at" maybe not the greatest idea?

On Thu, Feb 4, 2016 at 3:57 PM, Alvaro Herrera wrote: > David G. Johnston wrote: > > > ​Learning by reading here... > > > > http://www.postgresql.org/docs/current/static/wal-internals.html > > """ > > ​After a checkpoint has been made and the log flushed, the

Re: [HACKERS] "using previous checkpoint record at" maybe not the greatest idea?

On 2016-02-03 09:28:24 -0500, Robert Haas wrote: > Would we still have some way of forcing the older checkpoint record to > be used if somebody wants to try to do that? I think currently the best way to force an arbitrary checkpoint to be used is creating a "custom" backup label. Not that nice.

Re: [HACKERS] "using previous checkpoint record at" maybe not the greatest idea?

On February 5, 2016 2:52:20 AM GMT+03:00, Jim Nasby wrote: >On 2/4/16 3:37 PM, Andres Freund wrote: >> On 2016-02-03 09:28:24 -0500, Robert Haas wrote: >>> Would we still have some way of forcing the older checkpoint record >to >>> be used if somebody wants to try to do

Re: [HACKERS] "using previous checkpoint record at" maybe not the greatest idea?

On 2/4/16 5:09 PM, David G. Johnston wrote: What the 2nd para in the documentation is saying is something different: it is talking about reading all the pg_xlog files (in reverse order), which is not pg_control, and see what checkpoint records are there, then figure out which

Re: [HACKERS] "using previous checkpoint record at" maybe not the greatest idea?

On 2/4/16 3:37 PM, Andres Freund wrote: On 2016-02-03 09:28:24 -0500, Robert Haas wrote: Would we still have some way of forcing the older checkpoint record to be used if somebody wants to try to do that? I think currently the best way to force an arbitrary checkpoint to be used is creating a

Re: [HACKERS] "using previous checkpoint record at" maybe not the greatest idea?

On Mon, Feb 1, 2016 at 6:58 PM, Andres Freund wrote: > currently if, when not in standby mode, we can't read a checkpoint > record, we automatically fall back to the previous checkpoint, and start > replay from there. > > Doing so without user intervention doesn't actually

[HACKERS] "using previous checkpoint record at" maybe not the greatest idea?

Hi, currently if, when not in standby mode, we can't read a checkpoint record, we automatically fall back to the previous checkpoint, and start replay from there. Doing so without user intervention doesn't actually seem like a good idea. While not super likely, it's entirely possible that doing

Re: [HACKERS] "using previous checkpoint record at" maybe not the greatest idea?

On 2016-02-01 17:29:39 -0700, David G. Johnston wrote: > ​Learning by reading here... > > http://www.postgresql.org/docs/current/static/wal-internals.html > """ > ​After a checkpoint has been made and the log flushed, the checkpoint's > position is saved in the file pg_control. Therefore, at the

Re: [HACKERS] "using previous checkpoint record at" maybe not the greatest idea?

On Mon, Feb 1, 2016 at 4:58 PM, Andres Freund wrote: > Hi, > > currently if, when not in standby mode, we can't read a checkpoint > record, we automatically fall back to the previous checkpoint, and start > replay from there. > > Doing so without user intervention doesn't

Re: [HACKERS] "using previous checkpoint record at" maybe not the greatest idea?

On Mon, Feb 1, 2016 at 5:48 PM, Andres Freund wrote: > On 2016-02-01 17:29:39 -0700, David G. Johnston wrote: > > ​Learning by reading here... > > > > http://www.postgresql.org/docs/current/static/wal-internals.html > > """ > > ​After a checkpoint has been made and the log