Resource icon

Redis View Counters 1.1.3

No permission to download
I think Redis need to be installed on the server to use this addon ?
Correct. Hence why Redis was listed as an additional requirement.

Certainly does. However, even with redis properly setup and the redis-cache addon configured I ran into a handful of errors. Might just be me...
What sort of problems? I've got this running on Spacebattles and Sufficient Velocity just fine, and the hard part is configuring your config.php file as required for the redis-cache addon.
 
Is there any drastic improvements using it?

As i may need to install that module on server...
 
Is there any drastic improvements using it?

As i may need to install that module on server...
It is a fairly minor improvement. I have Redis configured to periodically save content, and on restarting it will restore the contents on start.

Compared to an in-memory table in MySQL which is totally lost on restart, and introduces memory contention for other resources. XF inserts a row for each page view per thread, where as this addon increments a counter per thread.

If you have a distributed MySQL instance, this means you don't need to replicate a MEMORY table for what is used to collect page view counters.
 
Correct. Hence why Redis was listed as an additional requirement.


What sort of problems? I've got this running on Spacebattles and Sufficient Velocity just fine, and the hard part is configuring your config.php file as required for the redis-cache addon.

You're absolutely right. Editing my original comment as it was user error.
 
  • Like
Reactions: Xon
@Xon nice gives me a good excuse to play with redis more as I already use it for queued tracking in Piwik Web Analytics.

For this addon which redis database does it default to ? is there an option to specify which redis database number to use if i want to keep things separate ?
 
Last edited:
  • Like
Reactions: Xon
And it is "appendonly yes" ?
Yes. I also have fsync set to explicitly off (so the OS will do it eventually). The idea is for low overhead persistence of cached data but as the data can be easily regenerated, I don't mind if it is lost.
 
Top Bottom