soedinglab / MMseqs2

MMseqs2: ultra fast and sensitive search and clustering suite
https://mmseqs.com
GNU General Public License v3.0
1.39k stars 195 forks source link

"Can not fit databased into 65536 byte" #9

Closed jhcepas closed 7 years ago

jhcepas commented 7 years ago

I just compiled mmseqs2 from latest source in a AVX2 machine and got the following error:

**$ grep -c '>' test.fa**
50
$ mmseqs createdb test.fa test.db
Program call:
test.fa test.db

MMseqs Version:         7885b8e5b2d4cb016d1c84455ae7d35b728497bd
Max. sequence length    32000
Split Seq. by len       true
Use fasta header        false
Offset of numeric ids   0
Verbosity               3

Time for merging files: 0 h 0 m 0 s
Time for merging files: 0 h 0 m 0 s
**$ mmseqs createindex test.db**
Program call:
test.db

MMseqs Version:         7885b8e5b2d4cb016d1c84455ae7d35b728497bd
Sub Matrix              blosum62.out
K-mer size              0
Alphabet size           21
Max. sequence length    32000
Split DB                0
Spaced Kmer             1
Threads                 40
Verbosity               3

Substitution matrices...
Can not fit databased into 65536 byte. Please use a computer with more main memory.

same happened using the linux avx2 binaries... but not with the sse4.1 binaries

martin-steinegger commented 7 years ago

I guess its the new ulimit code I added. Could you execute the command "ulimit" on this machine and send me the output?

jhcepas commented 7 years ago
$ ulimit
unlimited
martin-steinegger commented 7 years ago

Getrlimit RLIMIT_MEMLOCK does not behave how I expected. I changed the code back. Can you test it again?

jhcepas commented 7 years ago

it works now, thanks!