Line: 1 to 1 | ||||||||
---|---|---|---|---|---|---|---|---|
| ||||||||
Changed: | ||||||||
< < | Prevayler is controvertial because of their hype-filled site, and their claims about how much faster they are than a relational database. In the end, it turns out that prevayler is designed for working with small datasets that can fit in memory. It's not as scalable as a relational database, nor is it as "open" since people can't hook up to it and send arbitrary SQL commands. It's nice solution for cases where an RDBMS is overkill for working with a small data set. | |||||||
> > | Prevayler is controversial because of their hype-filled site, and their claims about how much faster they are than a relational database. In the end, it turns out that prevayler is designed for working with small datasets that can fit in memory. It's not as scalable as a relational database, nor is it as "open" since people can't hook up to it and send arbitrary SQL commands. It's nice solution for cases where an RDBMS is overkill for working with a small data set. | |||||||
Prevayler's javadoc blows. Why don't open source hackers document their code? | ||||||||
Line: 10 to 10 | ||||||||
Prevayler (see PrevaylerImpl constructor) first tries to recover the dataset from a snapshot (see the snapshot manager) and then recovers transactions from the journal. It does this by adding itself (an inner class) as a subscriber to the publisher. The publisher calls the journal which reads and replays any transactions that are in the journal since the checkpoint. | ||||||||
Added: | ||||||||
> > | TransactionSingle-node case
| |||||||
Snapshots |