3 ms·

I mentioned something similar in a separate comment. My initial DB schema was pretty bad. We did at 2 schema rewrites and migrations from the launch to 5M user

I mentioned something similar in a separate comment. My initial DB schema was pretty bad. We did at 2 schema rewrites and migrations from the launch to 5M users. Each time it took 2 weeks of sleep less nights. The machines today are really powerful. You can do a lot with 244 GB RAM machines backed by SSD. Someone who doesn't have the skill set to be able to scale once they get traction - it's likely they will not have the skills to design for scale at start. My recommendation to everyone would be pick a language and db you are most comfortable with and get started as soon as you can. You will fail on the product side a lot more times before you will fail on the technical side. And if you are failing on technical side, reach out to me. I will definitely be able to help you find a way out. I am not sure if there is any product guy in the world who an make a similar claim on the product front. However there are at least dozens of technical guys in the world who can make a claim like I did. So focus on launching the product as soon as possible. Work hard, reach out for help if needed. You will eventually get success.

developer2 10y ago
>> You can do a lot with 244 GB RAM machines Is this a typo of "244 GB" instead of "24 GB"? Nearly any company that has a single machine provisioned with 244 GB of RAM is doing something severely wrong, likely putting the company's ability to grow at risk. Such a machine screams of trying to vertically scale a poorly performing legacy product instead of figuring out to horizontally scale out with 16-64 GB servers. That much memory on a single server is a huge red flag for 95-99%+ of companies. It takes a very specialized system (ie: you probably don't fit the mold, no matter what your excuses are) to require such a server.
- bpicolo 10y ago
  Pretty sure it's not a typo. Amazon i3.16xlarge is 488 GB of Ram, 64 vCPUs, and a 20 Gigabit connection. Postgres/Mysql can absolutely scream perf-wise on that.
  - timewarrior 10y ago
    +1 - it wasn't a typo. DB machines benefit a lot from huge RAM.
  - fulafel 10y ago
    AWS also has 2 TB instances: https://aws.amazon.com/ec2/instance-types/x1/ https://aws.amazon.com/ec2/instance-types/x1/ Outside AWS, you can put 3-6 TB in a normal-ish server from Dell or HP, or 64 TB in a big iron box from Fujitsu or IBM.
- ne01 10y ago
  I know Basecamp uses a server with 2TB of RAM for their single MySQL server. FYI CTO of Basecamp is the creator of Ruby on Rails -- smart people. Checkout stack overflow stack too. Their system is faster too.
  - developer2 10y ago
    That is actually insane. The day their MySQL cluster experiences a severe downtime incident, it is likely going to take an enormous amount of time to recover. That is my intuition. If you know of a video or blog post (conference, talk, article, etc.) where someone explains the benefits of a 2 TB MySQL server and how it is not a crazy bad idea, I would love to see it. Because my 15 years of professional experience screams "NO WAY IN HELL" at that one.
- fulafel 10y ago
  Keeping all data cached in RAM is something that eats a bunch of memory, but is not usually wrong.

developer2 10y ago

>> You can do a lot with 244 GB RAM machines Is this a typo of "244 GB" instead of "24 GB"? Nearly any company that has a single machine provisioned with 244 GB of RAM is doing something severely wrong, likely putting the company's ability to grow at risk. Such a machine screams of trying to vertically scale a poorly performing legacy product instead of figuring out to horizontally scale out with 16-64 GB servers. That much memory on a single server is a huge red flag for 95-99%+ of companies. It takes a very specialized system (ie: you probably don't fit the mold, no matter what your excuses are) to require such a server.

bpicolo 10y ago
Pretty sure it's not a typo. Amazon i3.16xlarge is 488 GB of Ram, 64 vCPUs, and a 20 Gigabit connection. Postgres/Mysql can absolutely scream perf-wise on that.
- timewarrior 10y ago
  +1 - it wasn't a typo. DB machines benefit a lot from huge RAM.
- fulafel 10y ago
  AWS also has 2 TB instances: https://aws.amazon.com/ec2/instance-types/x1/ https://aws.amazon.com/ec2/instance-types/x1/ Outside AWS, you can put 3-6 TB in a normal-ish server from Dell or HP, or 64 TB in a big iron box from Fujitsu or IBM.
ne01 10y ago
I know Basecamp uses a server with 2TB of RAM for their single MySQL server. FYI CTO of Basecamp is the creator of Ruby on Rails -- smart people. Checkout stack overflow stack too. Their system is faster too.
- developer2 10y ago
  That is actually insane. The day their MySQL cluster experiences a severe downtime incident, it is likely going to take an enormous amount of time to recover. That is my intuition. If you know of a video or blog post (conference, talk, article, etc.) where someone explains the benefits of a 2 TB MySQL server and how it is not a crazy bad idea, I would love to see it. Because my 15 years of professional experience screams "NO WAY IN HELL" at that one.
fulafel 10y ago
Keeping all data cached in RAM is something that eats a bunch of memory, but is not usually wrong.