Skip to content

Instantly share code, notes, and snippets.

@lazywithclass
Last active March 21, 2017 04:15
Show Gist options
  • Save lazywithclass/55b21a847e802048f6ee01016fe11509 to your computer and use it in GitHub Desktop.
Save lazywithclass/55b21a847e802048f6ee01016fe11509 to your computer and use it in GitHub Desktop.
[CSE373] Lecture, winston-cloudwatch process inspection, Ubuntu 16.10 daily crash (-168)

[CSE373] Lecture, winston-cloudwatch process inspection, Ubuntu 16.10 daily crash (-168)

CSE373 2012 - Lecture 18 - Exhaustive Search & Backtracking

https://www.youtube.com/watch?v=usxrn66JDEo&feature=youtu.be&list=PLOtl7M3yp-DV69F32zdK7YJcNXpTunF2b

winston-cloudwatch process inspection

Started creating a couple scripts to mimic a real life process that logs to AWS CloudWatch, basically a running process that listens for SIGUSR1 and a sh script that sends that signal.

The goal is to find out if calling setInterval in winston-cloudwatch is working as I think it is, I am seeing some weird behaviour. One example is that logEvents is passed through to the integration library file as empty, which is something that should not happen, feels like I a missing something obvious about how Node.js works in this scenario.

Also it looks like when I'm logging twice in a row as in

winston.loggers.get('category1').error('1');
winston.loggers.get('category2').error('2');

two intervals are started, which is something that should not happen.

Dealing with continuous crashes on my Ubuntu 16.10

I have half a dozen error reports a day, plus once a day the screen freezes with a gray color with a graphical effect as if it was an old TV being turned off (a central white line collapsing to a dot).

Sometimes I get an extended error report saying

avahi-demon crashed with SIGABRT in avahi_malloc

This is avahi-demon man page https://linux.die.net/man/8/avahi-daemon, which looks like a service discovery tool.

I gave a look at avahi-demon --help and it looks like avahi-demon -k killed it.

I will give it a couple days to see if I still get so many crashes and error reports. If it will turn out to be responsible I could either remove it, update it, or just disable it at startup, so if I will ever need to discover something like a printer I could always turn it back on. (and of course I will never remember of this decision)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment