Unexpected move of VM while node crashed

Discussion in 'Post-Deployment: Configuration and Setup' started by Machiasiaweb, Aug 23, 2016.

  1. Machiasiaweb

    Machiasiaweb Mega Poster

    Messages:
    178
    Hello,

    I have 3 nodes running with same hardware and Virtuozzo license config. One of node are free without any VPS/VM running inside.

    Following is SHAMAN setting:

    # shaman -c cluster get-config
    LOCK_TIMEOUT=60
    POOL_CHECK_TIMEOUT=30
    LEASE_CHECK_TIMEOUT_FOR_MASTER=10
    LEASE_CHECK_TIMEOUT_FOR_SLAVE=10
    WATCHDOG_TIMEOUT=120
    WATCHDOG_ACTION=netfilter, reboot
    RESOURCE_RELOCATION_MODE=drs,round-robin
    LEASE_LOST_ACTION=reboot
    RELOCATION_SKIP_THRESHOLD=3
    CLUSTER_MOUNTPOINT_DEAD_ACTION=reboot
    VERSION=1

    Bot nodes are running 6.0.11-3475 now

    It just experience that one of the node crashed. However, those VPS/VM was moved to the another node which already have may VPS/VM there and it did not have enough license to allow to move. And the free node just migrated with those 'disabled' VPS.

    Does it any further setup I should config to allow VPS/VM migrated to free node next time?

    Thanks!
     
  2. SteveITS

    SteveITS Tera Poster

    Messages:
    253
    Usually there is a 72 hour grace period for containers/VMs moved to a new node. Our license (pstorage -c clustername view-license) shows "graceperiod=259200."
     
  3. Machiasiaweb

    Machiasiaweb Mega Poster

    Messages:
    178
    Oh. I expect it will move ASAP once accident happen. Does any way to change it.

    Now I have change RESOURCE_RELOCATION_MODE to SPARE
     

Share This Page