Redis report page runs out of memory

Our patch resolves a critical issue in Redis log page, previously broken when dealing with large Redis datasets. By adjusting the scan value from 10,000 to 1,000, the patch optimizes log retrieval, ensuring faster load times and improved usability.

Is the accuracy of data reduced when this parameter is reduced?

Technically, the patch changes a default value in ReportController::scan(). The reason why it works is that the controller in ReportController::overview() does not pass the second parameter. This leaves two places for a potential change. It would probably be better to set it in the overview method instead of relying on a default, IMHO.

The method is only used in a single place, I'm fine with adjusting the default. It's really just a wrapper to handle different implementations in the supported interfaces.

This bug was not fixed. After updating the Redis module to 1.8, we still get

PHP message: PHP Fatal error: Allowed memory size of 268435456 bytes exhausted (tried to allocate 20480 bytes) in contrib/redis/src/Controller/ReportController.php on line 129; PHP message: PHP Fatal error: Allowed memory size of 268435456 bytes exhausted (tried to allocate 20480 bytes)

every time the report page is open. Why is so much memory needed?

This isn't meant to fix all memory issues. The reports page attempts to provide a summary of possibly GB and millions of keys, it's entirely optional. There is another issue where I suggested a fixed limit or just completely skip parts of it of we detect a too large amount of keys. Profiling and patches welcome.

Automatically closed - issue fixed for 2 weeks with no activity.

The issue is not fully resolved. The page only works when the $count variable in the patch is set to 1. With any other value, the report page consistently triggers a PHP "out of memory" error.

This is not a global memory issue but specific to this page, as described earlier. Therefore, the issue should remain open until it is completely fixed.

I have reported this too but apparently

This isn't meant to fix all memory issues.

Create a new issue. The value is as tradeoff between memory and time. If only 1 works for you but otherwise you get an error then this seems like an indication that something very weird is going with your cache entries, they might be extremely large which would possibly also cause runtime issues outside of the report.

The issue I mentioned is ✨ Report page takes ages to load - maybe cache it? Active , but that won't help you.

When you create an issue, mention which client you're using, how much memory redis and PHP has and whatever else you think might be important.