> On Mar 29, 2019, at 6:58 AM, Michael Paquier <michael@xxxxxxxxxxx> wrote: > >> On Thu, Mar 28, 2019 at 09:53:16AM -0600, Rob Sargent wrote: >> This is pg10 so it's pg_wal. ls -ltr >> >> >> -rw-------. 1 postgres postgres 16777216 Mar 16 16:33 >> 0000000100000CEA000000B1 >> -rw-------. 1 postgres postgres 16777216 Mar 16 16:33 >> 0000000100000CEA000000B2 >> >> ... 217 more on through to ... >> >> -rw-------. 1 postgres postgres 16777216 Mar 16 17:01 >> 0000000100000CEA000000E8 >> -rw-------. 1 postgres postgres 16777216 Mar 16 17:01 >> 0000000100000CEA000000E9 >> -rw-------. 1 postgres postgres 16777216 Mar 28 09:46 >> 0000000100000CEA0000000E > Today there are 210 Mar 16 WAL files versus the originally reported 271. I cannot at this point confirm the original count, other that to say I used “ls -ltr | grep ‘Mar 16’ | wc -l” to get both numbers. Is it interesting that the earliest files (by ls time stamp) are not lowest numerically? Those two file names end “0000B[12]” (written at 16:33) in a range across the directory from “00001A” through “0000E9”? “0000B0” was written at16:53 and “0000B3” was written at 16:54 Is there any analysis of the file names I could do which might shed any light on the issue?
