FTR, there is a similar bug 1611973 about postgresql not starting in a
cloud instance, but this is specific to cloud-init, not to LXC. How sure
are you that this is really due to the low FDs of pid 1, as opposed to
that bug? Do you have some useful logs to look at, like a journal
output from that
I tried again with "classic" LXC (not lxd), and there the FDs look
differently: In xenial they point to /dev/pts/4, in yakkety they point
to /dev/pts/1. But installing postgresql in both xenial and yakkety
works fine.
I also tried with lxc-start-ephemeral, but in that case the low FDs are
again
Thanks for the initial analysis! This is very helpful. I'm marking this
as fixed in yakkety and add a xenial task.
I tried to reproduce this. I created a standard xenial and yakkety
container:
lxc launch images:ubuntu/xenial/amd64 x1
lxc launch images:ubuntu/yakkety/amd64 y1
In both of them
3 matches
Mail list logo
