Paper get rejected despite positive recommendations from both reviewers

This is a perfectly possible result. The reviewers' recommendations are just that: recommendations. The conference's scientific committee has the ultimate say over whether a paper is accepted, and they do not always follow reviewers' recommendations.

There may be quite legitimate reasons for a paper to be rejected despite positive reviews. For example:

• There might be an abundance of high quality submissions on the particular topic you submitted on (even if the deadline was extended because they didn't get enough submissions in other areas).
• There still could be a scope issue. Yes, a paper that is clearly out of scope should be rejected before it goes to review. However, if the committee felt it was at the margins of acceptable scope, they might send it to review but ultimately reject it because the reviewers' comments weren't sufficient to overcome the scope problem.
• The conference reviewers were generous. It is hard to interpret the meaning of the reviewers' scores without seeing the distribution of all scores. It is possible that the reviewers were really "nice" and scores tended to be high for everyone. It is quite difficult for many people to give a negative decision when they know that someone else is going to see the result and be affected by it, so scores can tend toward the positive, and rejection recommendations may be rare.
• There was some kind of rule preventing them from accepting it (This should have led to rejection before review, but it could have been missed. Of course, they also should have told you the reason if this were the case).

You are correct, however, that this is a fairly unusual decision. I think you ought to follow up and ask for clarification. It would be helpful for you to know the reason for rejection, and there is always the chance that there was some kind of clerical error, as well.

As ff524 says in a comment, the conference seems to be of a dubious nature, and it might be better not to want to publish there at all. Here are some of the red flags, but note that it is entirely possible for a number of excellent conferences to trigger one or more of them. It's just that all of them together suggest that some background checking is called for.

"The paper was reporting an original findings by the authors based on their simulation. xxx. I reckon it is of sufficient quality for the conference."

"was reporting" and "an original findings" reveals extremely poor English, and poor English is highly correlated with crackpot publications. "reckon" is quite informal and has not much place in a review, in my opinion.

"Good paper and well written. Easy to understand. xxx."

Like the other review, this one does not sound genuine.

The conference should have not received too many papers. This is because the deadline for submission has just been extended to 10 Nov. In other words, they continue to accept submission of papers at this point.

The best well-established CS conferences (especially the main ones rather than the smaller conferences with narrow focus) usually receive too many papers and end up rejecting even papers they deem a good fit for the conference simply because there are too many excellent submissions, and they usually do not arbitrarily extend deadline for submission just like that either.

the deadline for submission was first on 31/10. On 1/11, they extended it to 10/11. I received a decision on 3/11

As ff524 said, this implies that they reviewed your paper within 3 days. It's not impossible but just another tiny bell goes off, since there are already some ringing.

They reject your paper despite reviewers recommending acceptance but seem to be looking for more.

If you exclude the possibilities mentioned by dan1111, then this would be a warning sign.

At the very least, if you want to publish at a good conference, you had better be careful and check this one out thoroughly.