Skip to content

Commit cd4b6af

Browse files
committed
Fix unnecessary padding in incremental backups
Commit 10e3226 added padding to incremental backups to ensure the block data is properly aligned. The code in sendFile() however failed to consider that the header may be a multiple of BLCKSZ and thus already aligned, adding a full BLCKSZ of unnecessary padding. Not only does this make the incremental file a bit larger, but the other places calculating the amount of padding did realize it's not needed and did not include it in the formula. This resulted in pg_basebackup getting confused while parsing the data stream, trying to access files with invalid filenames (e.g. with binary data etc.) and failing.
1 parent 8225c2f commit cd4b6af

File tree

1 file changed

+5
-4
lines changed

1 file changed

+5
-4
lines changed

src/backend/backup/basebackup.c

Lines changed: 5 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -1638,11 +1638,12 @@ sendFile(bbsink *sink, const char *readfilename, const char *tarfilename,
16381638

16391639
/*
16401640
* Add padding to align header to a multiple of BLCKSZ, but only if
1641-
* the incremental file has some blocks. If there are no blocks we
1642-
* don't want make the file unnecessarily large, as that might make
1643-
* some filesystem optimizations impossible.
1641+
* the incremental file has some blocks, and the alignment is actually
1642+
* needed (i.e. header is not already a multiple of BLCKSZ). If there
1643+
* are no blocks we don't want to make the file unnecessarily large,
1644+
* as that might make some filesystem optimizations impossible.
16441645
*/
1645-
if (num_incremental_blocks > 0)
1646+
if ((num_incremental_blocks > 0) && (header_bytes_done % BLCKSZ != 0))
16461647
{
16471648
paddinglen = (BLCKSZ - (header_bytes_done % BLCKSZ));
16481649

0 commit comments

Comments
 (0)