Timeline for How can I scale an embedded Java DB to support multiple processes or nodes?
Current License: CC BY-SA 4.0
12 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jul 30 at 9:17 | comment | added | Basilevs | Related: Effective way to discover nodes in peer-to-peer network | |
| Jul 30 at 9:16 | history | edited | Basilevs |
edited tags
|
|
| Jul 5 at 22:52 | answer | added | Arseni Mourzenko | timeline score: 2 | |
| Jul 5 at 16:34 | comment | added | Basilevs | @PhilipKendall only Bittorrent with DHT is sharded. The original protocol relied on a tracker and was exclusively centralized. | |
| Jul 5 at 11:04 | comment | added | Basilevs | @PhilipKendall Bittorrent supports only one index, which makes it more similar to filesystem than DB. | |
| Jul 5 at 10:34 | comment | added | Philip Kendall | @Basilevs if you squint (a lot), Bittorrent is a sharded database on dynamically allocated nodes. Admittedly one with failure rates and efficiency orders of magnitude worse than tradtional databases. | |
| Jul 5 at 10:11 | comment | added | Basilevs | I would launch a dedicated DBMS replicating node from within fat jar. The key here is to use a suitable DBMS, but do not know any would support both symmetric deployment and autoconfiguration. The greatest concern is sharding support. I beleive sharding is impossible in a scenario of dynamically allocated nodes. | |
| Jul 5 at 9:55 | comment | added | Basilevs |
rebalance or replicate its data this is ridiculously hard to do. You are effectively trying to implement a custom scaling DBMS. Do not.
|
|
| Jul 5 at 7:51 | history | edited | Doc Brown | CC BY-SA 4.0 |
moved the unnecessary "update" paragraph into the question.
|
| Jul 4 at 16:13 | history | edited | xybrek | CC BY-SA 4.0 |
added explanation
|
| Jul 4 at 15:37 | review | Close votes | |||
| Jul 19 at 3:03 | |||||
| Jul 4 at 14:34 | history | asked | xybrek | CC BY-SA 4.0 |