Vantage has acquired ec2instances.info(vantage.sh)
vantage.sh
Vantage has acquired ec2instances.info
https://www.vantage.sh/blog/vantage-has-acquired-ec2instances-info
59 comments
I made some similar project https://ec2.shop/ it even support a curl interface. My proudest achievement is spot instances pricing. You can use it to bid automatically spot price.
Perhaps also worth mentioning is https://cloudoptimizer.io which has resource pricing for many cloud providers (not just EC2) and also has a way to filter GPU instances.
(disclaimer: my passion project, new version coming soon)
(disclaimer: my passion project, new version coming soon)
I use this project weekly! Thank you for building this, it makes it 1000x easier to estimate and plan projects.
This is the sort of thing I'd have acquired if I'd known it was on the market. Things like this never appear on the usual marketplaces for side projects in my experience, so do such sales tend to come from cold reach out?
What are the usual marketplaces for side projects?
MicroAcquire, Flippa, Indiemaker, and SideProjectors are worth looking at.
Why would you have bought it and what would you think its worth?
Would be awesome to add some kind of benchmarking, and cost per said benchmark value. Making it easier to evaluate which instances to pick over others on perf/$.
Yeah ended up doing the exact same thing with a csv export->Google sheets. Still not quite good enough I think since the different cpus/gpus perform differently.
That's a great suggestion!
I agree, vCPUs is really just an opaque proxy for CPU performance, having a CPU benchmark to sort by instead would be useful.
Please don't screw it up. I use this constantly.
Co-Founder of Vantage here: Priority #1 is not to screw it up.
As we mentioned in the blog post, nothing is changing at the moment but we are looking to add more functionality in time. Ultimately everything we do is driven by user feedback so please get in contact with us about how we may be able to help.
As we mentioned in the blog post, nothing is changing at the moment but we are looking to add more functionality in time. Ultimately everything we do is driven by user feedback so please get in contact with us about how we may be able to help.
As you're adding features, please maintain a high bar on speed and efficiency of the tool. I would really hate to see this thing bogged down with lots of fancy/unoptimized js libraries in the name of adding features
You will get a lot more SEO juice from that "Sponsored By" link by not moving it away from the original URL.
Good to hear!
As a user: PLEASE don't increas ethe amount of tracking and analytics. Twitter is enough. Remove it if you need something else.
As a user: PLEASE don't increas ethe amount of tracking and analytics. Twitter is enough. Remove it if you need something else.
This is a rule of thumb I’ve imposed on my product teams over the years. If you want to add (yet another) third party <script>, you have to be willing to drop one. It worked really well for us and forced product teams to focus harder on finding tools that tangibly help us as opposed to “looking cool”.
Thank you for the response. Regarding new features, I'd love to see Elasticsearch EC2 instance pricing added, to quickly get a ballpark month/year estimate.
Expansion of a similar format into other cloud providers and other AWS product offerings would definitely be useful.
Even if they do, it's on Github so anyone can fork away.
+1, it’s useful internally as a quick reference, too.
There's also an Azure version but it's been unavailable for me for some time - http://www.azureinstances.info/
Does anyone know if it's made by the same folks?
Does anyone know if it's made by the same folks?
Hey - ec2instances.info creator here. I only ran this for AWS and am not sure who may have run this Azure version, sorry!
This must be like one of a handful of .info domains that is not immediately bucketed into "spam" for me. Don't know why, but that's what that TLD means to me.
Creator here. Agreed! I don't recall how I decided this was the right domain at the time but I think the rare .info TLD made it easier for people to remember. Can't say I'll be using .info for future projects however...
Would love to know a ballpark figure for this acquisition. I mean this tool looks useful to some people I suppose - but the fact remains it's just a slightly more usable list of EC2 properties and costs. How much can it really be worth?
Not trying to sound snarky or look down on anything here. I'm genuinely curious. Cool tool. It's worth money!?
Not trying to sound snarky or look down on anything here. I'm genuinely curious. Cool tool. It's worth money!?
> it's just a slightly more usable list of EC2 properties and costs
For anyone who has to constantly (re)evalaute EC2 instances, this is a huge understatement. Now, I just looked at AWS EC2 pricing page and it looks way better than what I remember using in 2018.
Back then, for a given instance family, the specs and prices were on separate tables, making it maddening to compare. And forget comparing across instance families. instances.vantage.sh nee ec2instances.info made that so simple. Even the AWS ProServe employees that I worked with preferred it.
So AWS may have closed the gap but ec2instances.info is still superior since it offers advanced features such as "sorting" and "searching". The "Compare Selected" is super handy as well.
The biggest plus for Vantage is that ec2instances.info wired in to my fingers, I suspect that's true for a lot of AWS devs. By buying this site, they get a lot of eyeballs for their product. And frankly, it does make sense - Vantage is trying to selling you a better AWS console, this is a better pricing page, so it fits. The original dev gets some cash for a side project that was probably a huge headache to deal with. Win win for both parties.
For anyone who has to constantly (re)evalaute EC2 instances, this is a huge understatement. Now, I just looked at AWS EC2 pricing page and it looks way better than what I remember using in 2018.
Back then, for a given instance family, the specs and prices were on separate tables, making it maddening to compare. And forget comparing across instance families. instances.vantage.sh nee ec2instances.info made that so simple. Even the AWS ProServe employees that I worked with preferred it.
So AWS may have closed the gap but ec2instances.info is still superior since it offers advanced features such as "sorting" and "searching". The "Compare Selected" is super handy as well.
The biggest plus for Vantage is that ec2instances.info wired in to my fingers, I suspect that's true for a lot of AWS devs. By buying this site, they get a lot of eyeballs for their product. And frankly, it does make sense - Vantage is trying to selling you a better AWS console, this is a better pricing page, so it fits. The original dev gets some cash for a side project that was probably a huge headache to deal with. Win win for both parties.
Heck yeah it can be worth a lot in SEO value and reaching an audience of people looking at EC2 prices.
Total guess, but having some experience with transactions like this: $5-10k.
Making a mental note to track the SEO impact of the redirect to a subdomain here.
Total guess, but having some experience with transactions like this: $5-10k.
Making a mental note to track the SEO impact of the redirect to a subdomain here.
It’s worth enough to me to hope they don’t ruin it. I was using it every time I needed to check the info on some instance.
This has reminded me of a weekend project I made that compares AWS/Azure/etc in a similar table (https://providers.stacksetup.com/).
My wishlist:
* Could you fix the arch column, it currently shows "64-bit" for both Intel and ARM. No way to filter only ARM.
* Would it be possible to only show the latest generation?
* Would it be possible to hide disk/network variants? (e.g. C5D/C5N)
* Would it be possible to filter by CPU brand (Intel/AMD/Graviton)?
* Faster initial loading
* Quicker updates when AWS announces new instance types.
* Filter ranges (e.g. 4-8 cores)
* Dropdowns for filters (e.g. clicking on the arch will show all the options)
* Better datatable implementation (e.g. ag-Grid), to implement some of the above suggestions
* Could you fix the arch column, it currently shows "64-bit" for both Intel and ARM. No way to filter only ARM.
* Would it be possible to only show the latest generation?
* Would it be possible to hide disk/network variants? (e.g. C5D/C5N)
* Would it be possible to filter by CPU brand (Intel/AMD/Graviton)?
* Faster initial loading
* Quicker updates when AWS announces new instance types.
* Filter ranges (e.g. 4-8 cores)
* Dropdowns for filters (e.g. clicking on the arch will show all the options)
* Better datatable implementation (e.g. ag-Grid), to implement some of the above suggestions
These are great suggestions - thank you! I just copied it over to our company Slack and we'll see what we can do!
Ooh, while we're requesting columns - it would be great to add Core count (different from vCPU). See here for data: https://aws.amazon.com/ec2/physicalcores/
fantastic suggestions!
Mine: Option to do format with commas for thousands separator, only do dollars (no cents) for monthly and longer time frames.
Mine: Option to do format with commas for thousands separator, only do dollars (no cents) for monthly and longer time frames.
Acquiring and maintaining/improving web tools seems like a good way to spend marketing budget if you're a developer-focused business.
In the same category, there's a gcping.com testing latency between your browser and Google Cloud regions.
A Googler created that, but he's actually at Red Hat now.
More relevant though: gcpinstances.info exists and does what you think it should.
More relevant though: gcpinstances.info exists and does what you think it should.
I built https://runson.cloud to help you identify which Cloud or CDN a site runs on in the first place.
I would like to see one thing, approximate data transfer cost for instance with input field to write an amount to calculate.
Always found ec2instances.info super useful. But found vCPU such a weird unit (that's on AWS of course), as it changes with every CPU generation. So I bought cloudinstances.info. Originally wanted to demystify the vCPU. Never got around to actually building it.
Anything you would like to see?
Was thinking of the following:
- some stock synthetic benchmark
- response times of "todolist" app in django, express, ... for real-world data
- all kinds of io latencies and percentiles
- detailed CPU info including cache sizes (and maybe even cache thrashing of noisy neighbours as probably not everyone uses cache allocation technology yet)
- comparing with some dedicated hardware (would be interesting how nitro compares to kvm based cloud providers)
Any more ideas?
Anything you would like to see?
Was thinking of the following:
- some stock synthetic benchmark
- response times of "todolist" app in django, express, ... for real-world data
- all kinds of io latencies and percentiles
- detailed CPU info including cache sizes (and maybe even cache thrashing of noisy neighbours as probably not everyone uses cache allocation technology yet)
- comparing with some dedicated hardware (would be interesting how nitro compares to kvm based cloud providers)
Any more ideas?
vCPUs are just hardware threads (or "cpus" in Linux) which also change every generation; it has never been intended to be a measure of performance.
I remember raising an issue[1] back in 2015 asking for the unit to be changed from `cores` to `vCPUs`. Small world.
[1]: https://github.com/powdahound/ec2instances.info/issues/111
[1]: https://github.com/powdahound/ec2instances.info/issues/111
As a former contributor/co-maintainer and currently AWS employee I'm curious about your plans about the open source project behind the website.
Do you have any plans to change the licensing and/or contribution model? Are OpenSource contributions as welcomed as before?
Do you have any plans to change the licensing and/or contribution model? Are OpenSource contributions as welcomed as before?
We don't have any plans to change the licensing or contribution model. Open source contributions are still welcome as before. We aim to maintain the ethos of the project as much as possible.
The Github repo will be moving over to our organization this week and our employees will be looking at bolstering support moving forward.
I'm happy to answer anything further directly if you'd like -- just email me: [email protected]
The Github repo will be moving over to our organization this week and our employees will be looking at bolstering support moving forward.
I'm happy to answer anything further directly if you'd like -- just email me: [email protected]
As a current AWS employee, doesn't the popularity of this site prove the customer's demand for an official version to exist?
And then since it doesn't exist, Isn't it safe to assume that AWS intentionally obscures pricing?
And then since it doesn't exist, Isn't it safe to assume that AWS intentionally obscures pricing?
I think relatively few key people at AWS realize the level of popularity of this tool, but I know of multiple AWS employees who are big fans of it and a few have also contributed to it in the past.
AWS has been offering the APIs that the website uses currently as its main data source and also historically sponsored some of the hosting costs.
There's now also a similar table comparison view (although maybe not yet as convenient/comprehensive) available in the AWS console, but it's under active development and improving all the time. As far as I know the intent is to eventually have that view cover this use case. The AWS console team is eager to get feedback from customers to help improve it going forward.
AWS has been offering the APIs that the website uses currently as its main data source and also historically sponsored some of the hosting costs.
There's now also a similar table comparison view (although maybe not yet as convenient/comprehensive) available in the AWS console, but it's under active development and improving all the time. As far as I know the intent is to eventually have that view cover this use case. The AWS console team is eager to get feedback from customers to help improve it going forward.
A similar tool exists from AWS, the Instance Type (IIRC) inside the EC2 console.
As an aside, Garret Heaton is one of the co-founders of HipChat. Congrats on another exit mate!
Thanks! Initially created this when getting HipChat spun up on AWS. :)
There are a lot of good ideas in this thread, but recently I've found I can't use ec2instances.info anymore because there's no description of the instance classes (and I can't keep them all in my head anymore. c5 is obvious, but what are c5n, c5a, c5d, and c5ad? Might be troublesome in a column, but maybe a popover on the name column or something could work.
I am glad to know I am not the only one who cannot remember what the families and letters mean! There appears to be no slow down in sight either. I created https://cloudhw.info/ out of a desire to track these values.
This isn't bad at all, thanks for the link!
Still, with 11 general-purpose types for AWS alone I'm still lacking the context I need to determine which I delve into and what to disregard.
Still, with 11 general-purpose types for AWS alone I'm still lacking the context I need to determine which I delve into and what to disregard.
Hey, this is good feedback and likely a quick fix. If you email me ([email protected]) I can let you know when we have this support added.
That's super cool Ben, thanks! Email incoming!
There is a method to the madness:
-d is disk (instance store)
-a is amd
-ad is both
-n is enhanced networking (ENA, or EFA on the biggest instances)
-d is disk (instance store)
-a is amd
-ad is both
-n is enhanced networking (ENA, or EFA on the biggest instances)
Hey thanks that is simpler than expected!
Still, apart from the obvious ones (m, c, t) there's also the myriad top-level classes that I'd have no idea about - like what is r, i, d, x, p, g etc. My brain is too puny for all that!
Still, apart from the obvious ones (m, c, t) there's also the myriad top-level classes that I'd have no idea about - like what is r, i, d, x, p, g etc. My brain is too puny for all that!
These are my mnemonics, pretty sure I'm not too far off though:
https://stackoverflow.com/a/56880093
r = RAM. Instances are biased toward having more memory.
i = I/O. ... have fast SSDs.
d = DISK. ... lots of disk space.
x = XL. ... lots of RAM
g = GPU. ... have GPU hardware
https://stackoverflow.com/a/56880093
This is definitely helpful, though I think the SO question showcases fairly well why someone would want the summary on the web page: Several are ambiguous and if you're not regularly exploring new instance types you're going to forget this or fall behind in your awareness of new types.