Sickchill: scheduler "alive" statuses all red

[OneCDOnly]

Reported by @GTunney a few hours ago on the QNAP forum. This is a very recent clone from the develop branch :

How do we get these to go green again?

Thank you.

[issue-label-bot[bot]]

Issue-Label Bot is automatically applying the label Bug / Issue to this issue, with a confidence of 0.65. Please mark this comment with :thumbsup: or :thumbsdown: to give our bot feedback!

Links: app homepage, dashboard and code for this bot.

[miigotu]

Cannot reproduce this:

[miigotu]

Is that on python 3.8??

[OneCDOnly]

I'm not sure about @GTunney's version, but as it's installed via sherpa, it will be either Python 3.8.3 or Python 3.7.4

My Debian 10.5 installation running the same SickChill (from develop source branch) with Python 3.7.3 is also showing all red. :disappointed:

[miigotu]

Update from the develop branch and see if #6638 fixes it.

[OneCDOnly]

Update from the develop branch and see if #6638 fixes it.

Nope, no change: Still all red.

Not sure if this is relevant, but these services still seem to be running fine. The last run times for each are updated when they should be. :confused:

Also, the daily search and backlog search times are both shown as 0:00:00

[OneCDOnly]

update: just did a quick check by reverting back to a master branch install. All green alive statuses. daily search and backlog search times appear normal:

[OneCDOnly]

@miigotu what does the alive status actually mean? Might help track-down the issue.

[miigotu]

It's built into python, it's the function to check if a thread is running.

It doesn't make sense that it's not showing correctly if the threads are actually running.

[OneCDOnly]

OK, let me know if there's anything I can check that'll help.

[GTunney]

Sorry chaps only just morning here.

Yes I’m running python 3.7.4 via Sherpa.sh

[OneCDOnly]

Good work @miigotu - a fresh clone finally shows some green. :grin:

... although, not in the alive column yet.

[miigotu]

I'm gonna make a test app to see if it shows you dead threads in that. I'm feeling like y'all have a python that is patched by your distro or custom compiled with the incorrect options and that's causing the big, because the code is good on this side.

I will post it later.

[OneCDOnly]

I'm gonna make a test app to see if it shows you dead threads in that. I'm feeling like y'all have a python that is patched by your distro or custom compiled with the incorrect options and that's causing the big, because the code is good on this side.

Oh, then you're not going to like this. :wink:

I get the same problem on 2 different Linux platforms:

• Debian 10.5 with Python 3.7.3 from the Debian repos
• QTS 4.2.6 with Python 3.8.3 from OpenWRT

If I was to guess a common element between these two (other than I installed SickChill on both), it's that way-back-when, QTS was originally derived from Ubuntu (so, Debian after Canonical have fcsk'd with it). But the Pythons come from different sources.

As this worked on Python 2.7.16, maybe there's some extra module needed for SickChill to work in Python 3? (I know - not a lot of help, but my own Python experience is quite limited).

[miigotu]

It works fine here on Ubuntu Eoan in 3.7 and 3.8

[OneCDOnly]

Well, this is embarrassing. I tried a new install on my workstation - and installed everything in requirements.txt (which I hadn't been doing). Works fine. 👍🏽

My bad. 😞

@miigotu I'll address this in sherpa now, and my sincere apologies for stuffing you around.

[OneCDOnly]

BTW: this is what I see after having ensured everything in requirements.txt is installed:

[miigotu]

what wasnt installed? because it is all supposed to be in lib3 also

[OneCDOnly]

I hadn't specifically installed everything shown in the develop branch requirements.txt file. I was still working off my old list of pips that worked with the Python 2 version of SickChill.

After installing with pip3 install -r requirements.txt, the alive column works as expected. 😄

[miigotu]

lol

[OneCDOnly]

hmmm... looks like I'm not out of the woods yet. No joy on QNAP: https://github.com/OneCDOnly/sherpa/issues/76

That fix also didn't work on my production server (Debian 10.5), and it's now giving me a mako error on the show list page. I've just finished reinstalling Python 3 from scratch (no-small-feat), and it's still there. Cleared my browser cache too and tried another browser. 😞

Mako Error

A mako error has occurred.
If this happened during an update a simple page refresh may be the solution.
Mako errors that happen during updates may be a one time error if there were significant ui changes. 

May I ask: any idea if this error is something wrong with my config or database?

File /opt/SickChill/sickchill/gui/slick/views/inc_home_show_list.mako:206, in render_body:
data-src="${static_url(curShow.networknetwork_image_url)}"
AttributeError: 'TVShow' object has no attribute 'networknetwork_image_url'

That networknetwork almost looks like a typo.

[miigotu]

Sure looks like a typo to me.

[miigotu]

Fixed