Glitch project very unresponsive, gives sometimes errors

Hey @JaielSoft nothing that you and Melissa didn’t discuss - the “table-flipping” error is typical of problems between the browser and Glitch but not necessarily universally so. Network latency can be a cause. So can anti-virus or explicit network configuration issues. If you care to look at your network tab in your browser’s dev tools you might see a little more information about failed requests that we can explore.

As far as ping timeouts that sounds like a different issue. We haven’t seen reports of widespread issues and I’m not seeing anything in our monitors; can you share your monitor reports and basic configuration with us? Likewise with the project that you saw file corruption on - that’s an alarming thing to have happen and I’d like to get to them bottom of it. Is that on the project you previously mentioned?

no its this one: https://glitch.com/edit/#!/the-alpha-brian

I have the timeline for today. Its on Cula pinging every 2 minutes to my home page

Thats on all my project but they dont have those spikes at the same time though but on different times. It jsut got that bad the last few days.

There must have been a point where Glitch restarted my project because reading from the files only happens on startup and saving at finish. So it must have finished, corrupted the file and started up again reading from the backup file because the normal file was corrupted

I save the files as soon as a sigterm comes in (bot.js, commandmanager,js and filtermanager.js have a finish() command where they save the files, the invocation comes from init.js when a SIGTERM comes in), I guess something went wrong or maybe took too long to save the files but I doubt it 5 seconds should be enough for like 1mb max of data that needs to be written. But I have hourly backups so it wasnt that bad

EDIT: Upon further inspection, I believe there was no SIGTERM coming in but my bot’s process just got killed right away because the main file was not corrupted and the backup had the more uptodate data. The bot seems to never have had the chance to save the main file so either there was no SIGTERM or something else happened

@cori @mmcewen or someone else anything new on my topic. I dont know why my Bot just gets killed without the chance to save its state

@JaielSoft Can I be invited to the project, so I can take a look?
Send the project invite in a private message!

Hey so yeah. I got what was the error. My stupid mistake, was accessing an array element on a null/undefined reference without catching error. So I am very sorry to have wasted your time devs :sweat:

Ha well I’m glad you found it. I was composing a response indicating that I couldn’t find anything untoward happening (although there was one long stop time on 2019-03-04 (or it might be on 2019-03-05 - I’m trying to determine if the time I’m seeing is local or UTC).

Happy Glitching!

Aha, those times are local, so the long stop time was at around 0600 UTC on 2019-03-05.

stop time didnt see it though. gonna check on cula. nothing that I can see on Cula. Ive had some downtimes but those didnt exceed 2 minutes

sadsa