Обсуждение: Questions on pg_dump

We are doing some additional testing of our backup strategy with the goal of moving from the text dump to the custom format and have some questions about pg_dump -Fc.

Question 1.  Why would the custom format be larger?  pg_dump -Fc -Z 0 this morning was 88GB while the normal pg_dump last night was 82GB.  We have not had that much growth in the few intervening hours.  We were expecting the -Fc to be at the most the same size of the text dump, and really expecting the -Fc to be smaller.  Any explanations?  (We are doing timing with various levels of compression and with the use of external compression to see what gives us the best compression/time, so that is why the -Z 0 test.)

Question 2.  With the -Fc, how can we verify a successful backup?  With the normal pg_dump, we are examining the end of the file looking for:
--
-- PostgreSQL database dump complete
--

Is there any sort of token like that with the -Fc that can be used to verify a successful backup?

Thanks,

Chris

We understand that.  The question is why is an uncompressed Fc format so much larger than a straight dump?  Is the so bad it takes almost 5% extra space?