Make multiple attempts to startup rsyslog

[MartinFlores751]

Seems like rsyslog, after being killed, may either start up again by itself, or possibly start up in a delayed manner [1]. Making multiple attempts while also checking if rsyslog has already started up seems to make do for now.

---

[1] Previous observation when I ran into this issue. Might be incorrect or irrelevant.

[alanking]

Nice. Let's get some comments in here and I think it'll be good to go

[korydraughn]

Changes look fine to me. If/Once this is merged, we should capture the TODO in an issue (not before it's merged though).

This is just personal preference so feel free to ignore ... Consider removing the square brackets from around the attempt counter. It's a little shocking to see #[1] instead of #1.

Let's give @alanking a chance to think about this before adding the pound.

[MartinFlores751]

Fair on that, it was kind of an "Oh hey, we use '[]' when we throw in variables" sort of thing. Perhaps [#1] would be better if we want to keep that sort of style?

[MartinFlores751]

Also, should we mention issue #163 in the commit message? Or is that a separate thing?

[korydraughn]

Fair on that, it was kind of an "Oh hey, we use '[]' when we throw in variables" sort of thing. Perhaps [#1] would be better if we want to keep that sort of style?

Sure. That works too.

Also, should we mention issue #163 in the commit message? Or is that a separate thing?

Yes. That is exactly the thing we are providing a temporary fix for.

I'm hoping the RHEL9 people or Rsyslog people provide the true solution in the near future.

[alanking]

Looks good. Let's resolve open conversations, squash to one commit, and consider linking to #163(?). Leave off # for now.

[alanking]

Looks good. Please add #!