GitHub Outage - How databases are managed in production

[@akashanand3783]

Thanks for this informational video

[@AsliEngineering]

Thank you so much Akash 🙌

[@manish-mk]

Love to see your consistency brother ! I always wait for your videos as well as newsletters :)

[@goku_ui_god]

Can you please provide the handwritten notes for all of your videos posted on youtube starting from System Design? It will be of great help for all of us.

[@dpm_07]

Really a super cool video and now I know "A lot to know a lot to discuss".

[@dhruvilshah9098]

Thank you, learned many things new in this video.

[@sritejaparimi6605]

Hi Arpit, There will be multiple instances of proxySQL right? And if there are multiple the max connection pool each instance can have is MAX_DB_CONNECTIONS/No. of instances? I am thinking if we want to add/remove instances the connection pool for each instance also changes and how are we going to make sure it gets updated for each instance?

[@suryakantasahoo9935]

Great content !

[@d4devotion]

Ok so now I got it how database are managed in production. But wondering how do you manage your time by being a data engineer manager in Unacademy and creating such quality contents. :D

[@AsliEngineering]

I wake up at 6 every day 😀 to either prepare for the video or shoot the video ⚡

[@adianimesh]

@@AsliEngineering kudos !

[@tharun8164]

Is single node recovery done manually? Does it necessarily require developer intervention?

[@anupampandey3758]

if they reverted, did it not cause other issues which needed them to upgrade proxySQL version

[@SudhanshuShekhar6151]

Awesome learning, Thanks again. I am afraid, being such a big company they must be having Release notes for any deployment, at least it helps the team to narrow down the issue during the crisis.

[@pushprajsingh595]

Is there an alternative like proxysql for mongdb.

[@tharun8164]

Wouldn't there be data loss due to automatic failover? since typically the replication will be async.

[@tharun8164]

Could you add related resources, text books to your videos for further deep dive. Thanks.

[@manmohanmundhraa3087]

May be restarting replica server then making it master will work. Other reason may be because replica is optimized for read not for write....

[@subee128]

Thanks

[@mayuragrawal4410]

Orchestrator while promoting a replica to be the master, shouldnt it also spawn a new replica to have same number of nodes configured initially???

[@shwethanayak2459]

flapping like a domino effect

[@tushargupta764]

Another leaning: if your reverts fail as well, keep reverting until you have no code left :p

[@AsliEngineering]

No code, no bug, no outage 🔥🔥

[@sauravshaw6965]

GodStuff