One big "gotcha" for AWS newbies which I cannot tell if this addresses: Does this set or allow the user to set a cost ceiling?
AWS have offered billing alerts since forever. They'll also occasionally refund unexpected expenses (one time thing). But they've never offered a hard "suspend my account" ceiling that a lot of people with limited budgets have asked for.
They claim this is a competitor for Digital Ocean, but with DO what they say they charge is what they actually charge. I'm already seeing looking through the FAQ various ways for this to exceed the supposed monthly charges listed on the homepage (and no way to stop that).
Why even offer a service like this if you cannot GUARANTEE that the $5 they say they charge is all you'll ever get charged? How is this different from AWS if a $5 VPS can cost $50, or $500?
That's what Amazon is missing. People want ironclad guarantees about how much this can cost under any and all circumstances. I'd welcome an account suspend instead of bill shock.
Via some accidental clicking in the control panel (trying to get an IP address for the instance, I think?) I ended up getting a bill from them for over $100. Which, to me at the time, was a huge amount of money.
It put me off of AWS forever. I don't ever want something that tells me how much they're going to charge me after I have already given them my credit card information.
edit: they did credit me back when I complained, but that doesn't matter. The risk to me wasn't/isn't worth it.
From FAQ:
> What do Lightsail static IPs cost?
> They're free in Lightsail, as long as you are using them! You don't pay for a static IP if it is attached to an instance. Public IPs are a scarce resource and Lightsail is committed to helping to use them efficiently, so we charge a small $0.005/hour fee for static IPs not attached to an instance for more than 1 hour.
The Zeno's paradox in action - once you reach half your limit, the speed is cut in half. "Zeno" throttling if you will. :)
It feels like the larger thing they're trying to solve, that I expect actually stops the majority of people who don't choose AWS, is the complexity around setting up VPCs/SecurityGroups/Subnets/etc.
Most providers in the VPS space already charge overages for bandwidth, and most of them don't support suspending the account vs just billing you.
So in theory (I haven't checked what the API gives you control over - so this may be worthless), you could monitor your instance (bandwidth, time up, disk usage, etc), and if things get out of hand, or approach your limit (whatever it is), you could use the API to say shut down or delete the instance, or throttle the bandwidth (maybe via firewall rules or something?).
Again - this would assume the API allows you to do this (and ideally from within the instance itself - which shouldn't be an issue, I wouldn't think). And again, it shouldn't take this much work (you're right, it should just be a simple control panel setting).
But maybe it's an option for those who have the skills to implement it?
Exactly! We have used AWS and DO a lot this past year. DO is great for smaller sites/api's and super easy to use. Their support is also outstanding.
AWS has tons of tools but many come at a cost. We are in the process of moving a couple of sites off of AWS and onto a LiquidWeb dedicated box. We will be paying much less and the LW dedicated server is more than enough for what we need.
AWS is great for spinning up and scaling instances quickly and comes with a ton of other tools. At the end of the day however it is not always the most cost effective or even best offering for most sites/apps.
>For every Lightsail plan you use, we charge you the fixed hourly price, up to the maximum monthly plan cost.
Wording implies the monthly pricing is a 'maximum' price.
The big question here is what to do with stateful data. Would you accept an immediate deletion of all of your S3 data? RDS instances and snapshots?
- Setup a HTTPS endpoint on the server that listens for an SNS notification and performs an action (e.g. backup ephemeral data to S3 and shutdown). I wrote the service in Go and the action is just a shell script but choose your favorite language.
- Setup an SNS subscription pointing to the service endpoint.
- Setup an SNS topic for the message.
- Set up an SNS notification in AWS billing. I use "When actual costs are equal to 100% of budgeted amount".
The problem is that it's necessary to lock down the endpoint listener as it will usually need root access in order to shutdown the machine. This can be done by using authentication on the endpoint, setting up a locked down user to run the service under and granting that user the ability to run /sbin/shutdown in the sudoers file.
There are probably nicer ways to do it, but this does work to limit my spend on each instance.
You can also add AWS API calls to delete any other costly related resources (static IPs, load balancers etc.)
I've been thinking about writing a more modular and robust app that handles multiple instances etc but most of my servers are now in GCE so I don't really have the need.
Some types of data transfer in excess of data transfer included in your plan is subject to overage charges. Please see the FAQ for details.
The only overage charge I see is for data transfer. This isn't ideal, I'll grant you, but it's not the same as "various".
That's probably where most of their profit comes from. That option was most likely squashed from the highest authority.
Then I get the next months invoice and it's not using my pre-paid services but instead is billing for full CPU usage - no reserved instances.
After emailing support several times they say it's my own fault for not using the correct instance type, even though it's identical to the one I pre-paid for. It may well be my error but it was caused by them since I never asked for my servers to be moved. It's been an expensive and time-wasting experience -- will never use them again.
Am currently evaluating GKE (even more expensive) and DigitalOcean.
says Someone1234. Should we believe it?
Of course all things are not equal (i.e. CPUs, SSDs, bandwidth, etc).
Provider: RAM, CPU Cores, Storage, Transfer
----------
$5/mo
LightSail: 512MB, 1, 20GB SSD, 1TB
DO: 512MB, 1, 20GB SSD, 1TB
VULTR: 768MB, 1, 15GB SSD, 1TB
----------
$10/mo
LightSail: 1GB, 1, 30GB SSD, 2TB
DO: 1GB, 1, 30GB SSD, 2TB
VULTR: 1GB, 1, 20GB SSD, 2TB
Linode: 2GB, 1, 24GB SSD, 2TB
----------
$20/mo
LightSail: 2GB, 1, 40GB SSD, 3TB
DO: 2GB, 2, 40GB SSD, 3TB
VULTR: 2GB, 2, 45GB SSD, 3TB
Linode: 4GB, 2, 48GB SSD, 3TB
----------
$40/mo
LightSail: 4GB, 2, 60GB SSD, 4TB
DO: 4GB, 2, 60GB SSD, 4TB
VULTR: 4GB, 4, 45GB SSD, 4TB
Linode: 8GB, 4, 96GB SSD, 4TB
----------
$80/mo
LightSail: 8GB, 2, 80GB SSD, 5TB
DO: 8GB, 4, 80GB SSD, 5TB
VULTR: 8GB, 6, 150GB SSD, 5TB
Linode: 12GB, 6, 192GB SSD, 8TB
----- Dear Vultr Customer,
Including pending charges, your account is carrying a $5.94 balance.
In order to cover your current balance and your estimated monthly costs, our billing system will automatically deposit $275.00 from your payment method on file in 24 hours.
EDIT: Anyone cares to explain his reasons behind a downvote?
What if I need a lot of CPU power, but not much bandwidth? What if I want lots of RAM, but don't need much disk space? What if I'd rather have an HDD with more storage than a faster SSD? There's nobody offering a "configure your own VPS specs" plan.
I have no idea why people think Amazon pricing is worth it.
I'd be grateful if you used my link https://vpsdime.com/aff.php?aff=1272
The docs appear to say you can add these to a VPC but I don't see how to do it.
They don't say the SSD storage is local, so I'm sure it's not.
A few runs with `fio` confirms this is EBS GP2 or slower:
The bench: "fio --name=randrw --ioengine=libaio --direct=1 --bs=4k --iodepth=64 --size=4G --rw=randrw --rwmixread=75 --gtod_reduce=1"
Lightsail $5:
read : io=3071.7MB, bw=9199.7KB/s, iops=2299, runt=341902msec
write: io=1024.4MB, bw=3067.1KB/s, iops=766, runt=341902msec
read : io=3071.7MB, bw=242700KB/s, iops=60674, runt= 12960msec
write: io=1024.4MB, bw=80935KB/s, iops=20233, runt= 12960msec
These appear to just be t2.nano instances (CPU perf is good, E5-2676 v3 @ 2.40 GHz, http://browser.primatelabs.com/geekbench3/8164581).
For advanced users, there isn't much compelling here to make up for the administration overhead. It's a little cheaper than a similar-spec t2.nano (roughly $4.75 on-demand + $3 for a 30GB SSD). The real win is egress cost; you can transfer EC2->Lightsail for free. 1TB of egress would be nearly $90 on EC2, but is only $5 on Lightsail.
In other news, EC2 egress pricing is obviously ridiculous.
All competitors seem to outstrip AWS on this. Do they have some legacy infrastructure that is just too big too upgrade to something more modern, or is this "on purpose"?
In total, I'm spending about $15,000/yr on AWS, and someone spending $5/mo gets their bandwidth 18x cheaper than me? Shouldn't it be the other way around, and I should be the one with the discount?
I get enough headaches dealing with reserved instances, and trying to buy them at the correct time of the year to line up with price drops. Now, I need to consider dumping my autoscaling groups, EC2 web servers, and moving them to LightSail? Why not just give us a fair price on bandwidth, instead of more complications?
> Data transfer OUT from a Lightsail instance to another Lightsail instance or AWS resource is also free while the private IP address of the instance is used.
It could even be worth it to set Lightsail up as reverse proxy and profit off of very cheap(for AWS) traffic e.g. for S3. I can't really believe they would allow this. Am I missing something?
DO is not great in this regard as their butcher their allocations, but Vultr gives each VPS a proper /64. Scaleway has partial IPv6 support (not for their bare metal could, but their VPS's do support it).
I urge you to vote with your wallets, unless you really like paying $1/month or more per IP for the foreseeable future.
https://aws.amazon.com/blogs/aws/new-ipv6-support-for-ec2-in...
2core/2GB/50GB is a better offering than DigitalOcean's. Sadly, it seems like they still don't have FreeBSD.
All have 3TB transfer
Linode(Best): 4GB RAM, 2 Core, 48 GB SSD
DO: 2GB RAM, 2 Core, 40 GB SSD
LightSail(Worst): 2GB RAM, 1 Core, 40 GB SSD
I use this exact instance on Linode for my site https://dictanote.co; that one extra core makes a lot of difference when you want to take a backup of db or something intensive like that.
- What kind of SSD?
- What kind of "core"?
- How much are you over-provisioned on the physical hardware?
vCPUs are even worse. Buy a 20-core Xeon box, split it across 60 tenants, each with 2 vCPUs each, or via 40 tenants with 1 vCPU?
Ow.
Lightsail should be a entry-level product to upsell more AWS stuff.
Same thing with OVH, you can get a powerful dedicated server with unlimited bandwidth for the price of a medium EC2 instance. I hope they will open their datacenter in California soon.
Keep in mind that in France, very few people know that Amazon is more than a website selling books and random stuff. But in the US, very few people know about OVH and Scaleway.
I have their 3 euro arm server and have no problem pushing more than a 100 megabit a second from it.
The major reason to use a VPS host instead of AWS is that AWS is complicated. This seems to be just as simple as DO or a million other VPS hosts, with the added benefit that it's easy to hook up to Amazon's other services if you need to.
The major reason to use a VPS host instead of AWS is the order of magnitude price difference. That’s not going away with LightSail either.
Look at all the French/German/Dutch hosters, they’re 10-20 times cheaper than AWS, and you get far better storage performance and cheaper traffic.
So essentially you're getting $90 worth of egress traffic for $5. It's even more obvious now that EC2 egress pricing is ridiculous.
What I love about VPSes as opposed to AWS, Azure, or Google is that you get a completely a la carte box with a direct interface right to the Internet and both IPv4 and an IPv6 /64. You can instantly provision "servers" that you can do anything you want with -- you can treat them like "pets" to run a personal blog or a legacy app, or you can herd them like "cattle" with your favorite management and provisioning tools. The pricing is great and the infrastructure is mix and match.
Many VPS providers (Vultr and I think DO as well) will even let you upload and install an ISO directly onto the KVM instance over the web. That means you can install OpenBSD or even weird OSes. I've heard of people putting wacky stuff like OS/2 in the cloud this way. Some even allow nested virtualization.
A VPS is ideal for a large number of common work loads, but not all. For things where I want to make extensive use of AWS's managed services or where I want to have something more akin to a private data center, EC2 and similar offerings from Microsoft and Google are great. But for those I want the whole enchilada. If I'm going there I want everything the EC2 management console and API gives me including full-blown VPC, etc.
This seems to occupy an uncanny valley. Without IPv6, direct networking, etc. it's a crummy VPS, but it's not as rich as EC2. The only pluses I see are direct access to AWS services (but if I want that I probably want EC2) and AWS's security and uptime "guarantees."
Problem with the latter is that it's largely marketing. I've routinely clocked 300-day-plus uptimes on Digital Ocean and I've also had EC2 instances mysterious die or go into a coma. They might have something to say on security, but I've never seen any real proof that AWS security is intrinsically superior to their competition. Neither DO nor Vultr has had a recent major breach AFIAK and they all seem to use the same virtualization tech.
If you're familiar with AWS then you can get a similar offering directly, particularly using reserved instance pricing.
It's nicely packaged their existing products (EC2, VPC, ...). So you can get Digital Ocean like experience on AWS. You can still tune the underlying services.
The thing that keeps me away from AWS services is the depth of the service - I need to be an expert in AWS on top of knowing how to configure my servers, which for now is maybe a non starter.
It does show you the power of packaging: with a simple domain and thrown together marketing page, you too can target another market segment.
This is an intriguing move and one that I'm sure DigitalOcean, Linode, and Vultr have been fearing may happen.
The pricing is on par with these alternatives in the VPS space.
Linode offers the double of the RAM for the same price for $10, $20 and $40 VPSs.
It'd be nice if they bumped the $20 plan to 3GB RAM (or go full Linode at 4GB)
[1] https://amazonlightsail.com/images/hero/index-6316203d.jpg
Linode and OVH, while not as prestigious as AWS and DO, offer much more fair pricing when you need more resources.
But, as others note, the variable cost factor seems to still be a sticking point. I can setup a Digital Ocean droplet, or Linode, or one of a dozen other low-cost VPS providers, for $5 or $10 a month, and I know it will never cost more than that. Maybe I'll bump into memory, disk, or bandwidth limits...but, AWS is a killer if you aren't careful. I used to maintain (and pay for, out of pocket) a non-profit's website on AWS, and the price ballooned while I wasn't paying attention, due to automated backups to S3 and some other stuff, and by the time I noticed was costing me $183/month, for a website that could easily run on a cheap VPS. My fault for not paying closer attention, not setting up cost alerts, etc., but I moved the site off of AWS and onto one of my own web servers, where it literally costs me single digit dollars to run (it has many GBs of email but otherwise is a small site with very low traffic).
So...unless they're giving me some reason to think I won't end up with a massive bill one month because of a popular post, or something, I probably still won't think "I know, I'll use AWS!", unless it's a situation where I need the scaling capabilities of AWS.
Then I used digitalocean because of the free 1 year server time github gave me and everything was a breeze. They had tutorials for a lot of stuff, like how pubkeys and privatekeys worked, how to use ssh, how a server works, how to use nginx/apache, and even node.js stuff. I got up and running quickly even though it was my first time using a VPS. It was super easy, and the best part was with my knowledge gained from DigitalOcean, I was able to start using AWS with relative ease.
I think Lightsail is a good competitor to DigitalOcean, good for newbies who can't exactly figure out how much their server will use and charge them. But imo, with the same stats and stuff as DigitalOcean as a newbie I'd stick with DigitalOcean just because of how helpful their tutorials are in general and how helpful their interface is.
~ speedtest-cli
Retrieving speedtest.net configuration...
Testing from DigitalOcean (192.241.229.48)...
Retrieving speedtest.net server list...
Selecting best server based on ping...
Hosted by Monkey Brains (San Francisco, CA) [5.93 km]: 2.132 ms
Testing download
Download: 921.09 Mbit/s
Testing upload
Upload: 705.31 Mbit/s
Network speed should generally not be the issue with AWS, it's disk iops where the non-local SSDs will make a major impact.
Unlike VMs, which statically allocate mem whether you use it or not, containers have the chance to grow and reduce mem as workfloads go up and down, which means you could pay for GB/h of mem usage on the right-sized # of vCPU base.
Not sure if any IaaS/PaaS is doing this.
Joyent pricing[1] is still for static resource allocation and not cheap compared to these larger players.
[0] https://www.joyent.com/triton [1] https://www.joyent.com/pricing
https://gist.github.com/xfalcox/3b99beac4935fd154a4cbeb540dc...
Random thoughts from shameless noob:
- I like Digital Ocean's OS/app images for speed / smaller projects. Looks like Lightsail offers this with Bitnami. Not sure how complicated that is in their console.
- Amazon IAM comes with a bit too much overhead for noobies like myself. Glad this doesn't require you to set that up when you just want a quick and dirty VPS.
- Bah. Resizing/upgrading requires you to do it through API currently. That kind of sucks.
- Nice! 30 day free trial.
Lightsail 2: $10/mo 1 GB memory, 1 core, 30 GB SSD, bandwidth cap -- Scaleway: $10/mo 8 GB memory, 6 cores, 200 GB SSD, unmetered bandwidth.
OVH virtual servers have also always been cheaper than Lightsail, and bandwidth is of course unmetered: https://www.ovh.com/us/vps/vps-ssd.xml
Now if they could just get to the point where $5 gets you a running Docker container running on the equivalent of the Lightsail VPS (without setting up the backing EC2 infrastructure like ECS), I suspect that's closer to the platform that many users really want to have ...
I don't have huge hopes for their business going forward.
I also wouldn't be too surprised to see some people using these as middle-man boxes to reduce transfer costs associated with EC2 - $5 for 1TB is darned cheap for AWS. Using one of these to back up data from some EC2 hosts would be a win.
Doesn't expose the full, intimidating complexity of the AWS management console and workflow.
Does anyone have experience with how "clean" the AWS IP addresses are? I'd hate to switch to Lightsail and have to deal with deliverability and spam blacklist issues. I've been fortunate to have had zero issues on Linode.
That said, you have to ask them to remove their port 25 throttle and set rDNS: https://aws.amazon.com/forms/ec2-email-limit-rdns-request
It used to be Linode had the better of everything comparing to DO. Faster CPU even if they are same core count. SSD had much faster IOPS. Less oversold therefore bandwidth were good. And there is no point giving you 100TB transfer per month if you are limited to 10Mbps port speed. Linode ran on shared 40Gbps Port and peak bandwidth were great ( for its price ). Then there is the quality of the Network, ping time between different ISP and Exchanges. Linode has consistently been better then DO. And not it offer double the memory.
But many are worried about Linode's security issues and therefore would not even touch them with a ten foot pole.
I have yet to see quality VPS that offer a whole package better then Linode. Vultr, OVH, Online.net and Scaleway included.
I am hoping Lightsail bring some competition here.
Lightsail looks excellent since the setup is just a gazillion times more user friendly than the standard EC2. A single page affair, a launch script, authentication, it's all there. Once launched, I get all the info and metrics I need.
I kinda wish they streamlined their usual console to this level, but this way it's fine as well. I don't tend to use S3 and EC2 as much as I'd like given its non-existent UX, but this gives me hope that Amazon is taking user experience seriously.
Sure, it may be underpowered compared to DO or Linode, but having all services under one roof is worth it to me. I'm happy.
* First, this is great. The simplified interface vs EC2 is terrific. This is the direction EC2 (and RDS and S3 and basically everything) needs to be going.
* Instances you start in LightSail don't show up in your EC2 console. I would expect there to be some kind of data sharing there.
* Similarly, creating a "static IP address" doesn't show up in your elastic IP list. I'm not sure if this is intentional, but to manage two different views of products that you're billing me for is... troublesome.
* Last, if I could migrate elastic IPs from EC2 to LightSail I'd be migrating all of my instances immediately. The bandwidth savings are massive. (Related: when is the 2TB limit for a t2.small going to be migrated over to EC2?)
I don't understand why everybody on this thread is complaining about overages? If you use more than the allocated amount you pay overages, simple idea. Why in the world, would you want your server to just shut down when you reach a limit?
The in-browser SSH also deals with the problem of students who are on Windows machine. You haven't hated PuTTY until you've tried walking a student through it on their Surface Pro.
+ You can use private networking for cross-instance communication.
+ You can even communicate with other AWS services by using VPC peering.
+ Hourly billing.
- The firewall rules are not shared. You can't create a single rule and attach to multiple instances.
? What are the network throughput caps? Can't find it anywhere?
? Are the disk physical or using EBS?
? Can you snapshot a running instance, or does it have to stopped?The main advantage of Amazon LightSail over DigitalOcean are: built-in firewall (instead of messing with iptables), managed database with AWS RDS, and using S3 with a low latency and no networking cost.
It can be a very powerful model if done well, or done poorly cost a lot for less performance.
I use Hetzner myself for my large projects but the machines were configured using the same scripts I practiced over and over again on DO.
1). support more linux distros (hint: Debian)
2). let you place instances in other AWS regions
3). let you pool bandwidth quotas across instances
4). improve the cpu/memory competitiveness of their $20+ plans
Also, don't like the firewall being configured via the dashboard (reminds me of the same crappy approach used by scaleway). Alarming too that their SSH console auto logs you in (even if you started an instance with your own public key rather than Amazon's).
Is there a service, that you know of, that would simplify AWS so that we can just use it with predictable expenses and ability to grow? Maybe I am asking for too much.
Imagine if you could easily spin up DB instances and create LBs within LightSail - they would offer more features than DigitalOcean (or any other "VPS Provider"), while being price and usability competitive.
I would definitely be nervous if I was DigitalOcean - there are still advantages (support, tutorials, etc), but this closed the gap significantly.
t2.nano = $5
t2.micro = $10
t2.small = $20
t2.medium = $40
t2.large = $80LightSail's default firewall opens ports 22 (SSH) and 80 (HTTP) but has 443 (HTTPS) closed. That seems like a terrible default for making a developer-friendly service. Hopefully they fix that and open 443 by default. Otherwise, a lot of wasted time is going to be spent by developers who have configured SSL on their sites and don't know why it isn't working.
LightSail feels very similar to DreamCompute that DreamHost launched, including the approach of only allowing SSH public key auth without any option of using password auth. So, they're intentionally leaving out some users with that approach.
For this I prefer https://www.nearlyfreespeech.net/
DO and LightSail are close (if not) exactly the same spec wise?
All that's I've read so far essentially states that.
granted with AWS you probably get access to other AWS products so there's that...
I guess I google virtual private servers
$5/mo per 1TB of bandwidth = $5.00 / 1024 = $0.0049/GB compared to EC2's normal $0.09/GB -- That's a 91-95% discount on egress data!
I feel as excited as I was for Azure's Build 2016. Now I'm feeling pulled to AWS. This is great for AWS, not so much for Google Cloud which further fades into obscurity in my mind. I'd love to see that change, more competition in the cloud space = more options for us developers = more conditions in our favour.
Amazon LightSail just killed digitalocean for me which has been steadily getting more expensive (for instance I can't downsize to a less expensive plan once I resize my image, meaning I forked out $100/month for something that would work for $5/month + multiple DO images now cost monthly fee.
$5/month + tight integration with AWS products is enough for me to move completely off DO. If only AWS had DO's community style documentation, I'd definitely question DO's future viability.
Now a killer IDE from AWS that lets me deploy and configure AWS without leaving the IDE, that's a checkmate move, which I think will be very difficult for me to switch to another cloud provider. Right now things are in flux but I think an in-browser/desktop IDE like Cloud 9 with one click deploy to AWS would be the end game for other cloud providers.