How to compare two samples with significant different sample size?

I am trying to compare two groups with huge different size. Sample 1 is 34, Sample 2 is 11536. The data in both samples violate normal assumption and not equal variance. Should I do a random sample from sample 2, then compare Sample 1 and the sample of sample 1, or I should better just directly compare them. I am thinking of nonparametric test, like Mann-Whitney.

Thanks