join - Merge very large csv using pandas or awk -

i have 2 csv below (tried read them through pandas) df1 file 17gb (i read through pandas read_csv) , df2 700mb, want merge using trig_seq. python gets killed. there way through awk/join

>>> df1.head()    streamid  seqnum        timestamp_p1     trig_seq          1       1                14914503  10000000001          1       2                  1491450  10000000002          1       3                   1491450 10000000003          1       4                  1491450  10000000004          1       5                    149145  10000000005 >>> df2= pd.read_csv("/tmp/my.csv") >>> df2.head()      model_id  order               ctime     trig_seq          e62  1000000  1493311414272  30021182183          e62  1000001  149199641344  30021210134           e22   1000002  1491081210880  30021227875            e62   1000003  14951949824  30021239627            e62   1000004  14927136256  30021241522 >>> r1 = pd.merge(df1,df2) killed 

try -

$ cat f1          1       1                14914503  10000000001          1       2                  1491450  10000000002          1       3                   1491450 10000000003          1       4                  1491450  10000000004          1       5                    149145  10000000005 $ cat f2          e62  1000000  1493311414272  30021182183          e62  1000001  149199641344  30021210134           e22   1000002  1491081210880  30021227875            e62   1000003  14951949824  30021239627            e62   1000004  14927136256  30021241522 $ awk 'nr==fnr{a[nr]=$0;next} $2 in {print $0,a[$1]}' f2 f1|column -t 1  1  14914503  10000000001  e62  1000000  1493311414272  30021182183 1  2  1491450   10000000002  e62  1000000  1493311414272  30021182183 1  3  1491450   10000000003  e62  1000000  1493311414272  30021182183 1  4  1491450   10000000004  e62  1000000  1493311414272  30021182183 1  5  149145    10000000005  e62  1000000  1493311414272  30021182183