|
I am seeing the following logged in the apache error log (client IP removed):
That disk volume (/usr/local/apache/cache) used for the cache is an 80Gb tmpfs disk. The server has 32Gb physical RAM, 100Gb SWAP space. This is running on a RHEL6 server. At the point of the errors only about a quarter of the cache volume was being used: ~20Gb content. The server was reporting 29% memory used, with the remainder all being used for caching/buffering, and 6.5% (6.5Gb) of the swap space being used. My expectation of this configuration was: Apache would be able to cache up to 80Gb of content. Tmpfs would prioritise the most used stuff to be in RAM, and the least used stuff would be swapped out to disk (swap space). Based on this, all my memory / disk usage would seem to be ok. Is there something I'm missing, any other limits imposed by Apache / OS that might be affecting this? Any advice / help would be greatly appreciated. Thanks, Paul |
|