> I am working on parallelization MATLAB computing. Strangely enough, my > code sometimes works and sometimes fails, although untouched, with the > following error. > > Error using parallel.internal.pool.deserialize (line 9) > Bad version or endian-key > > Error in distcomp.remoteparfor/getCompleteIntervals (line 38) > origErr = parallel.internal.pool.deserialize(intervalError); > > Error in ga (line 68) > parfor idx = 1:numel(new_indi_idices) % only assess the necessary > > Caused by: > Error using distcompdeserialize > Bad version or endian-key > > I sorta believe this is a MATLAB bug, because it pops up very > irregularly. Sometimes it fails after running for 1 day +, whereas > sometimes it fails a few hours.
Hm, this is definitely not expected. Usually errors like this occur when the data transfer between the workers and the MATLAB client is truncated or corrupted in some way.
Do you have any simple self-contained example code that reproduces the problem? What cluster type are you using? What OS are you running on?
> P.S.: Suggested by "bad version", I tried both R2014a and R2013b, and > both versions have the problem.
Just to let you know - the "version" mentioned in the error is talking about the serialization version of the data, not the version of MATLAB.