[DRBD-user] Data corruption risk in protocol A,C.

Timofey Koolin timofey at koolin.ru
Fri Apr 24 09:56:53 CEST 2015

Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.


Is protocol A and C difference by data corruption risk?

I mean long time corruption when I have WAN replication with non ideal 
connections. Connection may be broken from time to time. Now I use 
protocol C. But I want to change it to protocol A.
I know about lost few last operations when I lost master node - it is OK.

I want to know - Has protocol A more data corruption risk then protocol 
C or not during replication. For example:
1. Write "123" to Master
2. Put data to tcp buffer in master, master return OK write data to OS.
3. Lost connection before data really send and accept by slave.

Will "123" put to slave after reconnect during resync process?



More information about the drbd-user mailing list