On 08/11/2015 01:38 PM, Jens Schmaler wrote: > On the other hand, this seems like a very common problem, so I was > wondering whether Condor offers some built-in magic to cope with this? > Maybe I am missing something obvious? Nothing that I know of. > Are there any recommended best practices for my case? In our case the problem is 100GB (and counting) of data files that change weekly. So far I tried scp, btsync/syncthing, lsyncd, ceph, gluster: they all suck. Some day I'll try data transfers as parent dag nodes on whole machine slots... If you have one .5G executable that doesn't change, and the number of execute nodes doesn't change too often, I'd just install it on every node and have some version of HAS_MY_SOFTWARE (look for it in the Subjects of recent list posts) in the classad. -- Dimitri Maziuk Programmer/sysadmin BioMagResBank, UW-Madison -- http://www.bmrb.wisc.edu
Attachment:
signature.asc
Description: OpenPGP digital signature