amazon web services - Need help regarding aurora DB scaling -


i need scale aurora db , down @ later sometime dont need downtime not bit. though of doing creating aurora replica , promoting scaling primary , promoting primary. involve downtime. please suggest alternative way.

i first ask in sense need scale up? writing? reading? if run server high-read/low-write ratio, i'd suggest read-replica (or more) , configure application use read replica in specific cases you're making queries , not modifying data. in sense, you're offloading writes separate instance of same data , letting primary deal writes. if did this, i'd suggest spreading out read replicas across other availability zones, in case az of primary goes down, read-replica auto-promoted primary.

armed info, suggest scale read replicas application configured use, can bring down replicas scale down without ever bringing down aurora primary. it's more of configuration problem in dealing multiple rds endpoints, essentially.

hth.

one caveat there delay in replication between primary , read replicas (my instances show 10-20 milliseconds), have keep in mind if perform write read in fast succession -- if first read after initial write happens 'too fast', replica might not see change , may either see no data (if you're creating) or stale data (if you're updating).

in general, isn't issue until you're under heavy load , writes on primary backed , start reading read replica before writes have been applied primary.

hth.


Comments

Popular posts from this blog

php - How to add and update images or image url in Volusion using Volusion API -

javascript - jQuery UI Splitter/Resizable for unlimited amount of columns -

javascript - IE9 error '$'is not defined -