Description
Make sure node failure detection do not use:
System.currentTimeMillis()
and
IgniteUtils.currentTimeMillis()
We should use nano time instead. Disadventages of current impl.:
1) System time has no quarantine of strict forward movement. System time can be adjusted, synchronized by NTP as example. This can lead to incorrect and negative delays.
2) IgniteUtils.currentTimeMillis() is granulated by 10ms
To fix:
ServerImpl.lastRingMsgReceivedTime
should be nano.
Attachments
Issue Links
- depends upon
-
IGNITE-13012 Fix failure detection timeout. Simplify node ping routine.
- Resolved
- links to