Here's what I know about DynamoDB
ฝัง
- เผยแพร่เมื่อ 1 ต.ค. 2024
- I didn't cover everything I know, but I just wanted to talk about some of the more important things you might need to know about Dynaodb.
📘 T3 Stack Tutorial: 1017897100294....
🤖 SaaS I'm Building: www.icongenera...
▶️ Generate Chapters: ytchaptersgene...
💬 Discord: / discord
🔔 Newsletter: newsletter.web...
📁 GitHub: github.com/web...
📺 Twitch: / webdevcody
🤖 Website: webdevcody.com
🐦 Twitter: / webdevcody
![ผัวเถื่อน - ยูริ โตเกียวมิวสิค [ OFFICIAL MUSIC VIDEO ]](http://i.ytimg.com/vi/P08ORRVDGl8/mqdefault.jpg)
This video convinced me to use Postgres 😂
This is highly appreciated.. would love more in depth videos like this one on AWS services!🙏 especially on how to approach things in production/practice like you highlighted with DynamoDB
Good job babe!!!
😘 love you!
🥺
the title should be "Why you should use relational database?" 😂
This makes me sad I built my react native app with amplify lol it’s okay these data models should last me long to migrate to an easier db
I have been working with DynamoDB from over a year now, will only recommended for Serverless + Smaller project where you already know what you want to implement (How you will fetching data etc)
Funny you recommend dynamodb for small projects when it's actually used for big and demanding projects.
@@everythinggoes850 by small project I mean the requirements, APIs, access patterns not the work load or traffic.
I actually ended up loving how dynamo db works. Since I’m about 1.5 years into coding with my startup and my first stack being react native and amplify I actually have no reference point of a prior DB. So the one table solution with PK, SKs, is turning out to be a joy to work with!
DynamoDB needs to much know how to use it effectively. There’s way better solutions put there
I haven't used DynamoDB and this got rid of a lot of misconceptions I had around it -- thank you 👍
17:50 "You get charged 1.25 million per write request unit" Damn, I do not think I can afford DynamoDB unfortunately.
Was that phrased correctly?
Hi, I have a question on a slightly different topic. If this topic has been discussed somewhere, I apologize. Namely, how do you organize your projects? I mean, structure and folders. One folder for all projects? Or many different folders? Division due to technology or something else? I ask because I always like the organization in your projects and code. It's very clean and everything has structure. Yes, I know that everyone has a different way and there are many possibilities, but I'm very curious what it looks like for you i.e. the organization of all local projects. And additionally, do you upload literally every project to the repository? Do you keep your most important personal projects on GitHub or, for example AWS CodeCommit?
I keep all projects on github. As far as project structure, I usually just colocate files as much as possible. Often I have one function per file if possible. I find it easier to manage vs having like 10 functions inside a single file.
have you tried go lang as a backend?
@WebDevCody What are the best practice to update items in a single-table design? What are the best practice for delete items or bulk delete? The case is for operation with items that could have duplicate attributes in different items.
I’m not sure what you’re asking. I’d personally use the update method whenever possible
@@WebDevCody For example, in SQL, you do UPDATE TABLE SET COLUMN = VALUE WHERE COL = COL_VALUE. It will update all the rows that matches the WHERE condition. DynamoDB doesn't provide any easy way to do update?
@@edpro4776 right you need to first query for all the records first then update them using batch updates.
Does Dax know you have DAX clusters?
shh don't tell him
in the nodejs code, where did u set up your aws credentials?
It’s set on env variables and the aws sdk automatically grabs those
You can also set up a credentials file. The docs for this are really good. There are several ways for doing this. After using AWS sdks for some years, I would suggest using what fits the project's needs. Let it be the credentials file, explicit declaration on the init of the ask or the environment variables. I think I used all of these approaches, in different projects. And sometimes combined, based on the environment where the project was running.
Great content, what is the name of the VSCode theme you use?
Bearded theme
Bearded theme
Can we use orms with dynamo db ?
Yes - Called AWS Amplify
in the video I mention Electro DB which is a cool library for interacting with dynamo
Friends dont let friends use Dynamo. Absolutely horrific dev experience. Dont do it. Once your data gets complex and you start needing rich data querying this db will kill you and you will have to do a major refactor. Learned this the hard way. Just don't do it!
yup, I basically say use SQL at the end
@@WebDevCodythe struggle is real my friend
True life story though, we built out entire SaaS against Dynanmo because of selling points and pricing structure on aws when compared to other solutions. Once we ran into the limitations I mentioned above. I was in a bad spot and took a major risk, used ChatGPT and ported the entire code base that referenced Dynamo to Mongo in a single weekend. And 95% of it worked out of the box. Almost making mongo a drop in replacement with way richer querying capabilities. All that did was buy us time. Soon we needed truly dynamic query building which forced us to move to Postgres. Expensive lessons learned….
I think we need to start understanding what we build more, before we pick up the technologies, especially for the DB. There are use cases where Dynamo DB fits really well. It is a key/value store, like redis, so don't use it as a relational DB and then expect to have the output of SQL DBs.
Just make sure dynamodb Will satisfty you data access pattern before using iy
I don't know the type of project you are working and client.
We do have technical which does understand the whole process of dybamodb otherwise working with non-technical client and using dynamodb will be terrible mistake.
Imagine client come up with simple requirements to have a filter by on data and you can't because of access patterns limitations. Don't wanna use multiple GSIs because of cost.