For users on Magento 2.3.7 through the latest version (2.4.3), we're seeing significantly elevated levels of outgoing data from the Redis instance being used for Magento's cache. The transfer usage often surpasses 1gbit/sec, but we've seen some sites spike as high as 14 gbit/sec. It is not only related to the CSP module as we have isolated that bug.
There appears to be an established bug report, but it doesn't appear to have gotten much attention: https://github.com/magento/magento2/issues/33685
Our team is seeing elevated outgoing data regardless of whether the site is using
backend to communicate with Redis. The issue seems to be unrelated to the backend being used.
For Magento instances that reside on a single server, there is a significant increase in outgoing data from Redis. However, this bug is more likely to be noticed by higher-traffic sites that have Redis residing on a separate server, as they're seeing bandwidth usage up significantly. With public clouds like AWS, that network usage can quickly add up to large billing increases, making this a bug that's costing Magento merchants a lot of money.
If anyone has any additional info on this bug (ie. If you're experiencing it, or if you've found a solution), please comment!
A few months ago I got the same issue, The reason of that issue is full page cache extension (third party), so when I have disabled it, the redis instance usage goes to normal state.
I would suggest you that, look for the extensions like full page cache and similar to that.
Hope this will help!
Thanks for weighing in, but looking at the conversation happening at https://github.com/magento/magento2/issues/33685, this bug appears to stem directly from recent versions of Magento 2, and not an extension.