Debian reconsiders init-system diversity

Re:Debian reconsiders init-system diversity by newnewpdro on Tuesday November 12, 2019 @02:22AM (#21512183)

Filtering journal contents by metadata is known to be unreliable.

It's not that the logs are missing, it's that they're not getting included by your query.

The way journald associates log lines with process metadata is inherently racy.

This situation has improved some with the metadata caching changes, but IIRC there are still some holes where processes can log without journald associating accurate/up-to-date metadata with the logged info.

[ Parent | on HN ]

Re:Debian reconsiders init-system diversity by pferde on Tuesday November 12, 2019 @03:12AM (#21512424)

That makes sense, I guess, in a rather warped and bizarre way. The point still stands - journald is unreliable, and after getting burnt like this it will be years before I start trusting it again.

[ Parent | on HN ]

Re:Debian reconsiders init-system diversity by newnewpdro on Tuesday November 12, 2019 @12:30PM (#21515988)

FYI, in case you wanted to document your pain, it might help prioritize fixing it fully:

https://github.com/systemd/systemd/issues/2913

[ Parent | on HN ]

Re:Debian reconsiders init-system diversity by dannyw on Tuesday November 12, 2019 @05:27AM (#21513029)

Great, so why don't we just keep text logs as text files?

[ Parent | on HN ]

Re:Debian reconsiders init-system diversity by mateuszf on Tuesday November 12, 2019 @06:08AM (#21513209)

Filtering text files in a structured way is even harder.

[ Parent | on HN ]

Re:Debian reconsiders init-system diversity by pmoriarty on Tuesday November 12, 2019 @10:18AM (#21514726)

Having unstructured text reliably be in log files is far better than having a broken log daemon forced upon you.

[ Parent | on HN ]

Re:Debian reconsiders init-system diversity by eecc on Wednesday November 20, 2019 @03:10AM (#21582017)

Dedicated tools exist for that...

[ Parent | on HN ]