If a person attaches a debugger to a Derecho process, and triggers a breakpoint, other members of the group that send messages or SST updates will retry a few times and then time out. The process being debugged will be kicked out as having failed.
It would be nice if people debugging an application could set a flag in the config file that would increase this timeout to a number like 2 minutes. Probably, on startup, we should print “Running in debug mode, failure sensing will be slow.” The user would then have 2 minutes to resume from the breakpoint before timeout occurs.
If a person attaches a debugger to a Derecho process, and triggers a breakpoint, other members of the group that send messages or SST updates will retry a few times and then time out. The process being debugged will be kicked out as having failed.
It would be nice if people debugging an application could set a flag in the config file that would increase this timeout to a number like 2 minutes. Probably, on startup, we should print “Running in debug mode, failure sensing will be slow.” The user would then have 2 minutes to resume from the breakpoint before timeout occurs.