What exactly do you need to get rid of?

page name
parameter name
some value of parameter

?

I want to deindex all parameter urls which are more than 1 million

I want to do very carefully which doesn't impact on site home page domain.com/index.html

only kill parameter urls through webmasters tool?

Have you read over this page? [support.google.com...]

ALL parameters? Go to the "parameters" section of wmt and tell google to ignore the parameters. They will probably already be listed.

Unless those same parameters do have meaning on other pages. Then you have to proceed more carefully.

You can indeed use a wild card in robots.txt something like you describe could be tested in wmt. But I would suggest to drop all parameters...

user-agent: *
disallow: /*?

Thanks Lucy & Jakebohall I have already set (No URLs) in URL parameters settings.

Now my question is:
(No URL) setting means Google won't crawl these urls anymore but will it deindex also? or I have to use url removal tool to drop these urls from google?

If i have to remove from url removal tool what url should I provide and will it impact my home page as well?

domain.com/index.html?action=*

(No URL) setting means Google won't crawl these urls anymore but will it deindex also?

Other way around. It may still crawl them if it finds them-- but results will be indexed together with the parameterless version.

Go to the "parameters" section of wmt and tell google to ignore the parameters

From my experience, using Parameters handling at WMT isn't much of use if you want to ensure Google doesn't count those URLs that carry specific parameters. It is just a directive, and Google may choose to act on its own volition. More often that not, I have seen Google neglecting the directive given.

You can indeed use a wild card in robots.txt something like you describe could be tested in wmt

It will only help in avoiding new URLs getting indexed, but the ones already in the index, will continue to be in the index.

tkchinese, as far as I know, finding a way to add robots noindex tag is the only workable solution, if that can be managed. Else, try your luck with parameters.

It will only help in avoiding new URLs getting indexed

Not even that: It will only help in avoiding new URLs getting crawled. It took me at least a year to wrap my brain around this fact, so I'm not letting go of it.

It will only help in avoiding new URLs getting crawled

:-) I will give it to you.

Friends I'm going to apply NoIndex...I didn't find any other solution.