On 23/11/16 16:31, Tom Lane wrote:
Robert Haas <robertmh...@gmail.com> writes:
[ Let's invent Oracle-style UNDO logs ]
I dunno. I remember being told years ago, by an ex-Oracle engineer,
that he thought our approach was better. I don't recall all the details
of the conversation but I think his key point was basically this:
- Reading a page that has been recently modified gets significantly
more expensive; it is necessary to read the associated UNDO entries
and do a bunch of calculation that is significantly more complex than
what is required today.
Also ROLLBACK becomes vastly more expensive than COMMIT (I can recall
many years ago when I used to be an Oracle DBA reading whole chapters of
novels waiting for failed batch jobs to roll back).
However I'd like to add that I agree this is worth looking at, as
ideally it would be great to be able to choose whether to have No-UNDO
or UNDO on a table by table basis...
Sent via pgsql-hackers mailing list (email@example.com)
To make changes to your subscription: