During BOINC Beta the assertion was made that applications were isolated from each other and thus one project's tasks could not interfere with another project's tasks. I pointed out then that since the two tasks had BOINC in common and knowing what I know about systems that there was linkage and time would eventually show tasks of one project interfering with tasks from other projects.
And I was right. The generalize "heartbeat" mechanism is such that tasks that have a long spin-up or spin-down time can cause the heartbeat to be lost and BOINC then kills all running tasks. And then later starts them up again, sadly, some tasks to not take to this behavior very well. If memory serves I was seeing this with IBERCIVIS, Hydrogen, and Drug Discovery on both OS-X and Windows. My point? Task and project isolation is more theoretical than real and even with increased isolation as Mark suggests will still not break the coupling through BOINC On Feb 23, 2010, at 7:55 AM, Mark Pottorff wrote: > So, if you are going to make changes, they should assure that all activity on > the client is well-behaved. This means that BOINC is not the only code that > should be insulated from any tampering by the applications. The applications > need insulation from each-other as well. Essentially you should have a unique > user per project. Otherwise one slot directory can clobber another. One > application to go up to another's project directory and trash all of the > stored files, etc. etc. _______________________________________________ boinc_dev mailing list [email protected] http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev To unsubscribe, visit the above URL and (near bottom of page) enter your email address.
