From what I have read if there are two WAEs at a site, and they are configured for hot failover, if one WAE fails, the second one immediately pick up compressing the data over the WAN link. I.E. no data flows over cold that has already populated the hash DB.
This is possible because both WAEs in this mode keep copies of each other's hash tables. Is this an accurate statement, and is the same true for rudundant central managers?
Secondly, after the failed WAE is repaired or replaced, does the WAS that has been carrying the load, transfter the hash DB back to the primary such that all data previously seen that is in the DB still transfers over warm?
Would this general example be true for both the WAEs and the Central Managers when configured for redundancy?