peak fragmentation ratio removed as it is a confusing field for users and trivial to compute at hand now that there is peak memory information in INFO output

save peak memory usage as statistic and show it in INFO. Also a new INFO field was added showing the fragmentation ratio using the peak memory info.

added minimal cluster section in INFO output. This is only useful to check if the instance is or not configured as a cluster node, all the other informations are accessible using the CLUSTER command.

DUMP implemented, RESTORE and MIGRATE fixed. Use zcalloc() instead of zmalloc()+memset() in sds.c when a new string is created with NULL initialization pointer.

centralized cluster config file name. Assigned slots in CLUSTER NODES output and in cluster config file.

propagate key eviction to slaves and AOF, like for expires, so that replicas and AOFs are always consistent even in maxmemory mode.

INFO commandstats section reset with config resetstat, a new microseconds per call field, and a few other minor things

Now INFO outputs the "INFO default" output if no argument is passed. Also added counts of calls for every command, thanks to an idea by Mike Shaver

Fixed issue #435 and at the same time introduced explicit ping in the master-slave channel that will detect a blocked master or a broken even if apparently connected TCP link.

fixed bgsave_in_progress in INFO when BGSAVEing with diskstore enabled, don't DEBUG FLUSHCACHE when bgsave is in progress.