Simon Deziel
Simon Deziel
> Do you see the issue only when the instance isn't running? A stopped instance stops being reported in the metrics. And to be clear, those errors did not occur...
They are always running, sometimes recently so (shortly after a host reboot) but often for a long while.
IMHO remote add token should expire too, people forget about things all the time. I myself found an unused token left over from a previous test. How about 24h? Or...
Would it be feasible to have a server config to disable or configure the delay before expiry? I'm surprised that cloud-like hosting are not too concerned with unused tokens. Maybe...
I know but that's however not always feasible to expose that socket to the consumer. Our use case is a prometheus exporter that runs containerized in a snap (https://github.com/optix2000/nsd_exporter/pull/8). This...
I just updated the PR to drop the now unused `data/Debian-8.yaml` and `data/Ubuntu-16.04.yaml` files.
I can confirm the problem as it caught us off-guard causing downtime :/ Commit 5d21b37a0617508c33fd4fd7c89f87c0ae4f2dea changed this area of the code and we now have to pass an empty hash...
It's on a zpool backed by a NVME partition. ``` $ lxc storage list +---------+--------+---------+-------------+---------+---------+ | NAME | DRIVER | SOURCE | DESCRIPTION | USED BY | STATE | +---------+--------+---------+-------------+---------+---------+...
Extremely fast outta be relative as the machine is from 2016 ;)
I'm happy to run any debug binary you feed me or give you SSH access (to my museum piece) if that's more convenient ;)