- Distributed Key value Database open source by payPal
- Earlier it was in single threaded c++ program similar to Redis, later written in go lang for better concurrency
- efficient cache also with persistance
- GoLang’s go-routines are light weight thread. It provides efficient posix thread than traditional thread which can multiplex over kernel thread.
- Limitation of Redis: Single threaded program will always run on one core, so its waste of resources if we have 4/32 core. but it do not impact redis because redis load is memory bound not cpu bound.
- JuneDB in other way can offer more heavy load computation as its a multithread key value storage.
- Provide 9.9999% of availability, means only down for 31.56 second in year
- Can handles 350 Billion request per day
Use cases:
- Provide temporary cache with TTL few seconds to few days(provide persistance as well).
- Idempotency: avoid duplicate process
Latency Bridging: Very fast inter cluster replication
