Quoting Christoph Mitasch ([email protected]): > Hello, > > I've tested the Nagios/Icinga Plugin check_lxc_memory.py. > https://github.com/bmiklautz/lxc-memory-utils/blob/master/check_lxc_memory.py > > I found out that the cgroup file memory.usage_in_bytes also takes the used > page cache into account. > Also the memory.failcnt file reports a failcnt as soon as the sum of RSS and > page cache memory exhausts the memory.limit_in_bytes. > Therefore the failcnt doesn't make too much sense for me at the moment. > > What is the recommended way to monitor memory usage of a container? > > memory.stat would give information about used page_cache, so that could be a > possibility. > https://www.kernel.org/doc/Documentation/cgroups/memory.txt > > Any news on /proc FUSE/virtualization?
That probably won't happen. Someone was going to start writing a library which could be used to get this information "from all the right sources" so that users could all use the common library instead of re-inventing the wheel. _______________________________________________ lxc-users mailing list [email protected] http://lists.linuxcontainers.org/listinfo/lxc-users
