Martin Sivák has posted comments on this change. Change subject: Added affinity Group collision test to canDoAction ......................................................................
Patch Set 3: (3 comments) https://gerrit.ovirt.org/#/c/41805/3/backend/manager/modules/bll/src/main/java/org/ovirt/engine/core/bll/scheduling/commands/AffinityGroupCRUDCommand.java File backend/manager/modules/bll/src/main/java/org/ovirt/engine/core/bll/scheduling/commands/AffinityGroupCRUDCommand.java: Line 71: /** I think there is a more readable algorithm eating about the same amount of memory: # Compute a list of linked VMs for each VM uag = Map<VM, Set<VM>> # Initial UAG from affinity groups for each positive hard affinity group pos_ag: for each vm in pos_ag: uag[vm] += pos_ag.allVms() # Transitional closure of the incidence matrix # (add all transitionally reachable VMs) do { changed = False for each vm in uag.keys(): for each other_vm in uag[vm]: uag[vm] += uag[other_vm] if uag[vm] size changed: changed = True } while (changed) # Check for positive vs. negative conflict for each negative hard affinity group neg_ag: for each vm in neg_ag: if neg_ag.allVms() intersection with uag[vm] > 1: fail Line 87: : Do you put all positive and negative groups to the uag here? What is the reason for doing that? Line 132: ag Why are you doing this after the conflict check? It has no influence on the result. -- To view, visit https://gerrit.ovirt.org/41805 To unsubscribe, visit https://gerrit.ovirt.org/settings Gerrit-MessageType: comment Gerrit-Change-Id: I56c483023ed14e238c8a67ec0890812158ecd0b2 Gerrit-PatchSet: 3 Gerrit-Project: ovirt-engine Gerrit-Branch: master Gerrit-Owner: Tomer Saban <tsa...@redhat.com> Gerrit-Reviewer: Gilad Chaplik <gchap...@redhat.com> Gerrit-Reviewer: Jenkins CI Gerrit-Reviewer: Martin Sivák <msi...@redhat.com> Gerrit-Reviewer: Roy Golan <rgo...@redhat.com> Gerrit-Reviewer: Tomer Saban <tsa...@redhat.com> Gerrit-Reviewer: automat...@ovirt.org Gerrit-HasComments: Yes _______________________________________________ Engine-patches mailing list Engine-patches@ovirt.org http://lists.ovirt.org/mailman/listinfo/engine-patches
