Problems in differential expression analysis with edgeR

I used glmQLF for differential expression analysis, and the result is almost all-down or all-up. Then I try glmQLF analysis with different data and glmLRT with the same data, but get the same result.
Here is my code:

group <- factor(c(1,2,1,2))
y <- DGEList(counts = counts_data,genes = rownames(counts_data),group = group)
isexpr <- rowSums(cpm(y) > 1) >=2
y <- y[isexpr,keep.lib.sizes=FALSE]
design <- model.matrix(~0+group)
y <- estimateDisp(y,design)
fit <- glmQLFit(y, design, robust=TRUE)
qlf = glmQLFTest(fit)
summary(decideTests(qlf))

result:

       group2
Down    12572
NotSig     13
Up          0

I think it may be something wrong with the design matrix because when I used exactTest which do not need a design for input, the problem seems to be solved. But I cant not find anything wrong with the design matrix. exactTest code:

group = factor(c(1,2,1,2))
y <- DGEList(counts = counts_data,genes = rownames(counts_data),group = group)
isexpr =rowSums(cpm(y) > 1) >=2
y = y[isexpr,keep.lib.sizes=FALSE]
y <- calcNormFactors(y)
y <- estimateDisp(y)
et <- exactTest(y)
summary(decideTests(et))

exactTest result:

         1+2
Down      31
NotSig 12512
Up        42

design:

> design
  group1 group2
1      1      0
2      0      1
3      1      0
4      0      1
attr(,"assign")
[1] 1 1
attr(,"contrasts")
attr(,"contrasts")$`group`
[1] "contr.treatment"