Thanks for the heads up. Just saw "gull" has issue with the build too. https://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=gull&dt=2018-11-09%2003%3A10%3A06
Both "gull" and "mereswine" run on the same machine. Probably wear out the SD card. Will stop these two animals and do a disk check. On Thu, Nov 8, 2018 at 7:36 PM Tom Lane <t...@sss.pgh.pa.us> wrote: > Thomas Munro <thomas.mu...@enterprisedb.com> writes: > > The last two runs show strange but identical (?) corruption in psql > > output on mereswine (Debian on ARMv7): > > > > https://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=mereswine&dt=2018-11-08%2009%3A10%3A14 > > > https://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=mereswine&dt=2018-11-07%2011%3A58%3A07 > > Note that the corruption is in the *expected* file, not the test result > file. I think this must mean that mereswine's git repo has become > corrupted. I'd recommend "rm -rf" on the repo and let the buildfarm > script pull down a fresh copy next time it runs. > > regards, tom lane >
