malhuda/jikan-rest
1
0
Fork 0
mirror of https://github.com/jikan-me/jikan-rest.git synced 2025-02-20 11:23:35 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

Update TROUBLESHOOTING.md

Browse Source
This commit is contained in:
Irfan 2019-08-18 22:21:27 +05:00 committed by GitHub
parent a53eecc1c2
commit 2987f4af32
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 24 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

31
TROUBLESHOOTING.md
View File

@ -2,26 +2,43 @@
## My local instance keeps returning HTTP 500
1. Check `storage/logs/lumen.log`.
First off, check `storage/logs/lumen.log`.
2a. If you're getting a Redis error along the lines of `BGSAVE` failing.
You're probably using `redis` as your `CACHE_DRIVER` (in `.env`). You probably have low RAM resource, switch over to file caching `CACHE_DRIVER=file`.
**1. If you're getting a lot of Redis errors along the lines of `BGSAVE` failing.**
You probably have a low RAM resource, switch over to file caching `CACHE_DRIVER=file`.
Redis throws this error even if you have about 1/4 of your RAM free. This is because it does a background save of your entire Redis cache - which is stored in-memory and which fails without the sufficient required memory.
You can stop Redis from haggling you and override that by running the following command: `redis-cli config set stop-writes-on-bgsave-error`
2b. Jikan is failing to cache (when `CACHE_DRIVER=file`) because the server is out of space - but you're sure you have enough space
Sorry! This is due to a bug on a previous release ([Issue #59](https://github.com/jikan-me/jikan-rest/issues/59)), please make sure you're upto date with on the releases
**2. Jikan is failing to cache (when `CACHE_DRIVER=file`) because the "disk is out of space" - but you're sure you have enough space available.**
Sorry! This is due to a bug on a previous release ([Issue #59](https://github.com/jikan-me/jikan-rest/issues/59)), please make sure you're upto date.
**Recovery Procedure**
- `sudo service supervisor stop`
- Delete Lumen logs: `rm storage/logs/lumen.log`
- Run `lsof | grep deleted` to check for the Lumen.log process, you'll know when it has a bizzare amount of space allocated to the process to it.
- Delete Lumen & worker logs: `rm storage/logs/lumen.log` + `rm storage/logs/worker.log`
- Run `lsof | grep deleted` to check for the "Lumen.log" process, you'll know when it has a bizzare amount of space allocated to the process to it. Copy the process ID and then kill it; `kill [process id]` e.g `kill 12345`
- Make sure your Jikan instance is on the latest release. Run a `git pull` and then `composer update`
- Reduce the number of supervisor processes in `/etc/supervisor/conf.d/jikan-worker.conf` e.g `numprocs=1`
And then reload the supervisor configuration:
`sudo service supervisor start`
`sudo supervisorctl reread`
`sudo supervisorctl reload`
`sudo supervisorctl update`
`sudo supervisorctl start jikan-worker:*`
- Restart everything:
`sudo service supervisor restart`
`sudo service apache2 restart`
`sudo service redis restart`
If Redis is taking too long to restart, follow this: https://stackoverflow.com/a/45069100/2326811 and then start it `sudo service redis start`