General linear model and residuals

Hi, i'm using SPSS for analysis of my data using general linear model with 3 independet factors (2 levels each) and one independent ordinal factor (4 levels). I saved the residuals and tested for normality with shapiro, but the residuals are no-normal. i'm reading in some blogs and some people say that residuals' normality it's not quite importante and can be ignore, but some people say that is really importante! i'm confuse. What should I do? I looked at some QQ plots and residuals seems to match with logistic distribution, but I don't know whether this means anything.