@grampelberg
Now, after making ksync handle only relevant events, this has issue got fixed, it's not crashing continuously 👍
We can take this PR live. Please check.

@alok87 it sounds like this doesn't fix it for you? Seems like a reasonable change either way.

@alok87 please sign your commits so that I can merge this.

@grampelberg There were 2 issues:

• First issue: Continous failure due to ERROR event object. This crash of sync, does not clean up the syncthing process which runs on :8384, so the next restart of ksync by our ksync wrapper(practl sync) also fails until the user manually kills the syncthing process. Handling only relevant events fixes it.
• Second issue is a repeated 1min40sec crash. I have opened a separate issue for it Ksync watcher should reestablish connection #337. Still finding the root cause.

have pushed the signed commit

@alok87 I'm trying to catch up here, so let me see if I'm understanding correctly. When we do an event request, and for some reason the cluster errors out, the returned event doesn't match type and the process fails. Is that right?

If I'm understanding that correctly this change might "fix" it in that it doesn't crash in this case, but it also won't catch any errors. Let me think about this a little bit and see if we can handle it gracefully.

@timfallmk out of the two issues i mentioned above

1. In the first (for the issue Ksync keeps crashing  #334) the process was exiting without clearning of syncthing process. But after this fix it is not a problem anymore.
2. In the second (Ksync watcher should reestablish connection #337), the exit is gracefull already, no dangling syncthing process is left behind.

I will give the PR for catching the error and logging it.

Got it @alok87 . I read it wrong. Looked like you weren't handling any errors from the events stream, but I see what's going on now. Merged and released in https://github.com/ksync/ksync/releases/tag/0.4.4

thanks @timfallmk