[mira_talk] Re: Mira3 server
- From: Bastien Chevreux <bach@xxxxxxxxxxxx>
- To: mira_talk@xxxxxxxxxxxxx
- Date: Fri, 27 May 2011 23:06:41 +0200
On Tuesday 24 May 2011 18:26:22 Juan Daniel Montenegro Cabrera wrote:
> that is the problem, the server and my laptop have both 4GB of memory and
> the estimated is 3.9 GB for de novo and 3.1GB for mapping and still both
> crash.
Remember that these numbers are just estimates.
Furthermore, do you run a 32 bit Linux? If yes, the process will not be able
to grow to the full 4 GB limit, but the OS will kill it as soon as it reaches
2.7 or 3.1 GiB.
> For some reason the log file is empty but the program loaded the sequences,
> qualities and trace info with no problems and started the assembly
> normally.
Errrmmmm .... if you did
mira ... --job=... ... >&log_assembly.txt
then the file "log_assembly.txt" *MUST* be on your computer, there simply is
no way it is not. Also, it should not be empty unless you ran out of disk
space from the very first nanosecond you started MIRA ... then you'd have
another problem though.
If the OS kills MIRA because of a bug, you should see something like
"Segmentation fault" on the command line (probably not in the log), if it was
killed because of memory constraints, you would see the very terse "Killed"
message.
The "log_assembly.txt" is where you will find the reason MIRA stops if it's
something internal.
B.
Other related posts:
