What exactly is OMA doing here?

Question

[moranr]

Hi,

Apologies for the title, I couldn't think of a more appropriate one.

Just trying to understand a part of the OMA pipeline , I'm using the -s flag and still on the multiprocessing part.

I currently have 1032 proccesors running with my OMA analysis. The nodes seem to be all working and efficiency is seemingly high, but all log files show something like below as the last line i.e. about to start a part but not yet starting

job_731906_task_6_of_1032.log <==
[pid  11158]: Computing genomeA vs genomeB (Part 3830 of 4200). Mem: 0.856GB

I would have expected to find some of the last lines to be 'mid process' between two parts like :

[pid  17178]: 95.00% complete, time left for this part=0.56h, 53.3% of AllAll done. Mem: 0.856GB

What exactly is going on when the last line of the log is like below?

Computing genomeA vs genomeB (Part 3830 of 4200). Mem: 0.856GB

Thanks

Answer 1

[Adrian Altenhoff]

Hi,

this is just the initial message before it starts chewing on computing the optimal pairwise alignment of the chunk (part) it mentions. afterwards, every 50k pairs, it will give an update status on stdout. Note that if the chunk has less than 50k pairs, noting will be printed before it prints another line mentioning that it finished with that part.

Depending on your cluster setup it could well be that the output gets buffered and is not yet printed.

Adrian

[moranr]

Great thanks for the information.