I have Amazon EC2 instance with NGINX + PHP-FPM (through socket) + APC.
I see sth. like this in iotop on high load.
Total DISK READ: 0.00 B/s | Total DISK WRITE: 0.00 B/s
TID PRIO USER DISK READ DISK WRITE SWAPIN IO> COMMAND
7200 be/4 www-data 0.00 B/s 135.12 K/s 0.00 % 0.00 % php-fpm: pool www
7202 be/4 www-data 0.00 B/s 308.84 K/s 0.00 % 0.00 % php-fpm: pool www
7183 be/4 www-data 0.00 B/s 61.77 K/s 0.00 % 0.00 % php-fpm: pool www
7185 be/4 www-data 0.00 B/s 69.49 K/s 0.00 % 0.00 % php-fpm: pool www
7186 be/4 www-data 0.00 B/s 73.35 K/s 0.00 % 0.00 % php-fpm: pool www
7189 be/4 www-data 0.00 B/s 196.89 K/s 0.00 % 0.00 % php-fpm: pool www
7191 be/4 www-data 0.00 B/s 158.28 K/s 0.00 % 0.00 % php-fpm: pool www
Most part of logs is disabled already, script doesn't write any files, but there is still lot of disk writes ( up to 500 IOPS according to AWS stats).
Is there any way to reduce disk writes?