On 2014-08-28 12:36, Konstantin wrote:
how would you pass 100000 id values to pilerexport?
Like i did for reindex
for ID in `cat /tmp/metadata.csv` ; do reindex -f $ID -t $ID ; done
It cound be a sinle id or range.
well, the ids are usually unrelated, so most of the time you just can't
specify a range in case of pilerexport. And forking pilerexport 100k
times gives you poor performance.
It may be better to read the ids (or id ranges) from a file, and
perhaps
it's also a time saver if pilerexport writes the messages to a single
zip file.
I'll implement it, and let you know if it's ready to test.
Janos
