SatNOGS Network/DB issue? bad gateway error when trying to vet BOBCAT-1 observations

I’ve tried vetting this BOBCAT-1 observation yesterday (still can’t now)

https://network.satnogs.org/observations/4833319/
It errors out after a couple minutes each time, I’ve tried several times.

Same thing again today from another BOBCAT-1 observation made less than an hour ago

https://network.satnogs.org/observations/4833317/

I get that black error page:

Maintenance in progress!

We will be back soon! Thank you for your patience.

For more info feel free to contact us.

502 - Bad Gateway

I can vet all my other observations without getting this error page, just not BOBCAT-1 with lots of frames. Not a big deal, just trying to keep up with the vetting process, obviously those two observations are good with ~2500 frames between them. Too many frames? :slight_smile:

1 Like

This is an known issue, indeed it comes from the many frames and it is related on the recent changes on the storage we use. I’m working on a fix for it, it should be deployed soon.

2 Likes

It looks like that the changes I did weren’t enough. So, it will need some more work and time for solving this.

2 Likes

Almost all spectrograms from my observations have dissapeared recently. What is going on???

https://network.satnogs.org/observations/?norad=&observer=&station=1260&start=&end=&page=2

it looks like a regression… checking it.

EDIT: check the progress of the fix in Fix regression that hides waterfall and payload (!1047) · Merge requests · librespacefoundation / SatNOGS / satnogs-network · GitLab

1 Like

The regression is now fixed… thanks @s59mz for reporting it!

2 Likes

It works now. Thank you for very fast response.

2 Likes

There is a temporary issue of not visible data in data tab of an observation, this is related with the recent fix for the issue in the first post. I’m working on fixing it.

EDIT: Here is the progress of the fix. The problem was that there was a different part of the code that runs for production site making the fix work on dev but not on production.

@satcolintel5 the issue is now fixed. Currently all the data are loaded in frontend, if the observation has more than 10 demodulated data then it loads the first 10 and give option to load (ten) more or all of the rest, which may take sometime if there are many data. Please let me know if you spot any issues on these changes.

1 Like

@fredy Thanks, that fixed it! The large frame count observations for bobcat-1 are opening instantly now.

Cheers

1 Like