Forum: Managing VoltDB

Post: problem about SnapshotDaemon

problem about SnapshotDaemon
zhangchengly
Jul 16, 2015
below is the log,

2015-07-16 03:30:08,981 INFO [SnapshotDaemon] LOGGING: Scheduling truncation request processing 10 seconds from now
2015-07-16 03:30:18,981 INFO [SnapshotDaemon] LOGGING: Snapshot truncation leader received snapshot truncation request
2015-07-16 03:30:18,991 INFO [SnapshotDaemon] LOGGING: Initiating @SnapshotSave for log truncation

"Scheduling truncation request processing 10 seconds from now" means what?

why it needs 10 seconds?

thanks.
pzhao
Jul 16, 2015
Zhangchengly,

Simply put, there is no problem with the logs you've provided. VoltDB is operating normally. VoltDB will wait 10 seconds before starting truncation snapshot. The 10 seconds is used to allow all nodes a chance to trigger their truncation snapshot requests.

Technically, truncation snapshot trigger is not coordinated across all nodes in a cluster. Thus, when a node triggers a truncation snapshot, it will give rest of the nodes a chance to address their truncation snapshot triggers during these 10 seconds. If a node triggers their truncation snapshot after 10 seconds, it will check the if a snapshot is in queue. If there is one, it will retry at a later time, otherwise it will trigger the same sequence of events.

Peter Zhao