Please close this bug. This box has been in service for a long time, with zero issues.
After a fresh install of stretch, this issue with perl exhibited itself. After, all seemed quite well. However, a week later? SSH connections started dropping with checksum errors. A memtest86+ run, of 2 *days* finally showed RAM errors. It wasn't until the 10th or 12th run through, that they appeared. This box had an uptime of -- well, a year prior to this upgrade. It seemed fine, and was a stable runner. The bad memory was very specific to a small part of one DIMM, so I can only assume that luck resulted in that memspace not being used for anything I noticed. Sorry for the issue guys, very sorry to waste your time. And thanks for the help. On Fri, 21 Apr 2017 18:00:25 -0400 Brad Barnett <debian-b...@l8r.net> wrote: > > > > > Can you reproduce the segfault reliably by running: > > > > "deb-systemd-helper unmask cgmanager.service >/dev/null" > > > > With Perl 5.24? > > > > Hmm. > > root@newoldbe:/var/cache/apt/archives# deb-systemd-helper unmask > cgmanager.service >/dev/null /usr/bin/deb-systemd-helper was not called > from dpkg. Exiting. > > No segfault. > > .... > > Now that my system is working, I just completed the apt-get -f, made > sure sysvinit was installed, removed systemd, purged it and such. > > I then did a purge of cgmanager and others (they were pulled in for some > reason with sysvinit), and tried a fresh install. It's fine. > > deb-systemd-helper still says the same as the above, with no segfault. > >
