3 ms·
Can anyone shed light on how persistence is achieved? The wiki page doesn't seem to come to any conclusions: https://ramcloud.stanford.edu/wiki/display/ramclou
by ricardobeat 14y ago
Can anyone shed light on how persistence is achieved? The wiki page doesn't seem to come to any conclusions:
https://ramcloud.stanford.edu/wiki/display/ramcloud/Data+persistence https://ramcloud.stanford.edu/wiki/display/ramcloud/Data+per...
- rektide 14y agoHire more server huggers, make extra sure das-blinkenlights stay blinking, buying a second backup generator. And- obviously enough- off site redundancy; multiple data centers. Which they talk about the difficulty of: 10Gbps pipes between DCs is not a lot compared to write throughput of a single computer's main memory, much less a DC's. For some compressible checkpoint logs there might be a way forward. Perhaps we need a new *-cast technology: after any-cast we'll all be needing n-cast, which sends our traffic to the n-nearest DC's? :)
- s_baby 14y agoIf it's a distributed network, data can be distributed across nodes using a RAID algorithm.
- rstutsman 14y agoHere's the long answer to your question: http://www.stanford.edu/~ouster/cgi-bin/papers/ramcloud-recovery.pdf http://www.stanford.edu/~ouster/cgi-bin/papers/ramcloud-reco...