- client: work fetch: if there are idle devices, we need to ask

the highest-prio project for work for all of them
    (don't scale by the fetchable resource share!).
    This should fix some device starvation problems.
This commit is contained in:
David Anderson 2013-02-11 13:19:39 -08:00 committed by Oliver Bock
parent af7948910e
commit c7a5156573
1 changed files with 1 additions and 3 deletions

View File

@ -456,9 +456,7 @@ void RSC_WORK_FETCH::set_request(PROJECT* p) {
}
}
// our share of the idle instances
//
req_instances = nidle_now * w.fetchable_share;
req_instances = nidle_now;
if (log_flags.work_fetch_debug) {
msg_printf(0, MSG_INFO,