Bulk checkpointing causes the master and backup run-time states to be synchronized. Since ExtremeXOS runs a series of applications, an application starts checkpointing only after all of the applications it depends on have transferred their run-time states to the backup node.
After one application completes bulk checkpointing, the next application proceeds with its bulk checkpointing.
If a failover occurs before bulk checkpointing is complete, the SummitStack reboots. However, once bulk checkpointing is complete, failover is possible without a SummitStack reboot.