P-value changes in the same sample set under the same analysis parameters

I was double checking my results and noticed that p-value for alpha-diversity was different when I performed the same analysis for the same sample set. This threw me off since I was expecting the same p-value every time. Is this supposed to happen?
Please let me know.
Thank you

Please follow our post guidelines.

The data submitted had the samples with less than 10k counts removed in data filter, with minimum count of 2, prevalence of 10% and low variance of 5% removed. Data was further rarefied to minimum library size and scaling (TSS) was applied. Alpha-diversity analysis was performed with default options (filtered data, feature-level, Chao1, T-test/Anova). Then, p-value was noted.
However, when repeting this again, with exactly the same steps, alpha-diversity p-value was different.
I repeated the same steps yet again and again alpha-diversity p-value was different those before.

I see you rarefied your data … which is basically a random subsample n from the total N items. The results are expected to be different especially for alpha diversity estimation. Please refer to this post for more discussions