I have around 50-60 Linux and Unix Boxes.All I want is to monitor the processes and memory for each machine.The Monitoring should be capable to retreive every single health of each boxes.With Single Health I did mean that it should have capability to monitor the process being utlized and if it goes beyond certain limit it should send alerts therein.
It should not just send alert but give possible solution for that too.
Any suggestion?

Search this site or use google.
This question comes up 4 or 5 times a week.

I googled it out but couldnt find the right tool.

You could always consider writing it. Though you couldn't find the right tool? There are several, what was wrong with the ones you tried?

I did explored Nagios and linux utility like top,sar etc. But what I need is detail information which can also recommend me the indetail understanding.

I agree with Jamescondron on writing your own if you cannot find one.

How about this idea:
You can put a little web server on each and every machine that spits out a cgi script which reads 'ps', 'uptime', 'top', etc, outputs which then is filtered to spit out what info you want when accessed by you. All these web pages can then be further accessed automatically by a monitor machine which you designate as the "Overseer" of sorts. They can then be accessed every umm... minute? or 5 minutes? I'm just guessing.

Idea 2: Maybe Webmin has such a feature? I used it before for
a single computer.
http://www.webmin.com/

But then again you want suggestions on what to do and that seems to me like a custom job.

checkout zabbix or spiceworks

Look at nagios again, then, possibly in conjunction with cacti ...