After updating to v0.13.0 /read will not load

[aphick]

Since updating the reader feed does not load. My reverse proxy sends a 504 and loads just hang on the debug message
[READ] Updating post cache

• Multi-user mode
• sqlite
• Open registration? no
• Federation enabled? yes

[thebaer]

Hi @stayinside, can you share some more information about your setup, as prompted in the original bug report template? For example, which database are you using?

And just to be sure, did you run writefreely db migrate after updating the binary?

[aphick]

Oh sorry my mistake!
Updated my post, and yes I definitely ran that.
Running with the debug flag does not add any more information

[DaarkMoon]

Same issue for me but there is a strange point:

• After a restart if go to go to my user page (https://writefrely.example.com/user) -> It's OK
• Then I go to the read page (https://writefrely.example.com/read) -> I get an http error (504 Gateway Time-out)
• If try to go back to my user page (https://writefrely.example.com/user) -> Now I have http error (504 Gateway Time-out)

Maybe the read page lock the sqlite database ?

Application configuration

• Single mode or Multi-user mode? Multi-user mode
• Database? sqlite
• Open registration? no
• Federation enabled? yes

Version or last commit: 0.13.0 (last binary released for linux amd64)

PS: I have run writefreely db migrate after updating the binary

[thebaer]

Thanks for the additional information @stayinside and @DaarkMoon -- I agree it sounds like it could be a read lock on the database. I'll try to replicate and debug more when I'm back from vacation this week. (I'm not sure what changes around the Reader query could've caused this -- it could be the additional GetCollectionAttribute call on each blog while looping through the overall Reader results.)

[bandicootdancing]

Describe the bug

Writefreely is hanging when clicking on the reader. I'm not sure if crawlers/users are triggering the reader but writefreely is crashing daily since upgrading. I'm not sure if it's related but by the time I get around to restarting writefreely I see "too many open files." I've disabled the reader to see if the instance stays up.

Steps to reproduce (if necessary)

Steps to reproduce the behavior:

1. Click on reader
2. See [READ] Updating post cache in terminal
3. Writefreely becomes unusable until restart

Application configuration

• Single mode or Multi-user mode? Multi
• Database? sqlite
• Open registration? no
• Federation enabled? yes

Version or last commit: 0.13.0

[poVoq]

I have a similar issue on a fresh 0.13 install (same setup as OP).

The site runs fine until I make a first post. Upon which it continues to run fine for 5 minutes or so, but the post never shows up in the /read part and soon after the entire site goes in permanent 504 timeout.

[thebaer]

Thanks @poVoq, this should be fixed in #477. If anyone else can test out that branch and verify the fix, that'd be much appreciated!

[DaarkMoon]

I have build binary fom branch fix-sqlite-reader-lock and now it's work for me.

[thebaer]

Great! Just merged that fix, and we'll release v0.13.1 later today.