Hi sekerob,

I run client 6.6.29 on MacOS X 10.5.7.

Nobody has finished it yet. Here is the name of the WU in question: CMD2_0007-ACTC.clustersOccur-CAH3.clustersOccur_3_16093_17035

Currently it is at 298.6% after 2:08 hours.

You can test if the job makes real progress by looking and the checkpointing. In BOINC Manager [v 6.6 and up], Task view, select the Task. On left there is the Properties button. Hit that, opening a new screen. It gives the last checkpoint written to disk and more info such as real CPU time used which normally should not be far behind elapsed [wallclock] time, if running BOINC at 100% of available spare time.

At 2 hours and by the suffix of the job 943 positions included, there should be fairly frequent checkpointing which you can also have recorded in the message log with a cc_config.xml file option.

edit: inserted version info for emphasis of feature availability!

Here is an other one with a strange behaviour
CMD2_0008-MYH2A.clustersOccur-MYH2A.clustersOccur_1253_1109484_1109560_1109495_1109503

When the wu had run for a litle more than 22 minutes and reched 22%
it jumped back to 11% and the runtime shows 11 minutes.

Thank you Lawrence!
But in that case BOINC display wrong runtime. The runtime should not jump back, just the estimater time left.
Is it posible that a wu can jump forth and back like this for days, weeks
or even months and still display a runtime of 11 minutes?

I will run 1 wu and se what happens.
Is a wu at 285% and counting also supposed to be normal whit this project?