Hello,
I got response from pgsql-hackers that it is normal. In fact, the
newer segments that you see are pre-allocated ones (none of those
contain any valid WAL data as of now). And WAL data is currently being
written to an earlier segment like 000000010000000E00000080 (which
seems to be the latest in your "ls -lrt" output).
By the way, what are the values of following configuration parameters
for your cluster?
checkpoint_segments=?
checkpoint_timeout=?
archive_mode=?
archive_command=?
--
Amit Langote