Uploaded image for project: 'BrownDog'
  1. BrownDog
  2. BD-841

elasticity module suspended more VMs than it should


    • Icon: Bug Bug
    • Resolution: Won't Fix
    • Icon: Normal Normal
    • None
    • None
    • None

      When scaling down, the elasticity module suspended more VMs than it should, so that the minimum # of extractor instances are not maintained. This issue was self-corrected in the next iteration that the module runs, when it found that the min # was not met and resumed a VM, so it's not severe. But it would be good to fix it.

      The reason is that we generated a candidate VM list to suspend based on their idle time, then check each one if suspending it can still maintain the min #. However, the checking does not take into account that, earlier in the same iteration and the same suspending loop, another VM could be just suspended, thus reducing the # of running instances of this extractor type.

              Unassigned Unassigned
              ruiliu Rui Liu
              0 Vote for this issue
              0 Start watching this issue