Dear mates, We have an Intel SSR212MMC2 system with 32GB of rams , 12 SATA HD;s (each 2 are made raid0) port 80 traffic is routed from a mikrotik to the squid 2.7 stable7 box (running on debian lenny) cache_dir aufs /cache1 1200000 16 256 cache_dir aufs /cache2 1200000 16 256 cache_dir aufs /cache3 1200000 16 256 cache_dir aufs /cache4 1200000 16 256 fdisk -l /dev/sdc1 1.8T 72G 1.7T 5% /cache1 /dev/sdd1 1.8T 72G 1.7T 5% /cache2 /dev/sde1 1.8T 72G 1.7T 5% /cache3 /dev/sdf1 1.8T 72G 1.7T 5% /cache4 we noticed that when cache is nearly about 280GB (70GB on each cache_dir), squid closes all sessions every 1 hour for 30 seconds, then it works back normally... So what can cause this problem? Iam sorry i didnt mention too much details, because iam afraid to flood you with data. Iam ready to provide any details you need. Cordially, Bilal
