High-Memory Instances and $5 Linodes | Hacker News Reader

High-Memory Instances and $5 Linodes

blog.linode.com

293 pointsjoshtronic170 comments

Linode is my top choice for hosting, but lately I've been tinkering on a project that needs a lot of disk, and it's hard to find good options there. On AWS every dimension scales independently, but with Linode I can't buy disk without buying CPU and RAM too. When I had a similar need a few years ago, I wound up renting a dedicated server, where even the cheapest came with 2TB or so. I'd love to just stick with Linode, but I need an option to add disk!

"In the coming months we’ll be releasing into beta our new Linode Manager, RESTful API, and block storage service."

From the blog post you just commented on ;-)

Good catch. Thanks!

I was in the same situation a year ago. Eventually found Vultr that allows attaching additional "disk instances". But despite that I would choose Linode for any resource-balanced project.

DigitalOcean has their new block storage https://www.digitalocean.com/products/storage/

Cool. Is "block storage" anything other than a fancy new word for "network drive"?

If you want disk, as in spinning rust, I think it's going to remain that way for a long time. For some reason cloud providers don't like to offer disk at a reasonable price. I suspect it might be because it's not actually a single disk but space on something like a RAID10 SAN, where costs are higher since it inherently requires more hardware.

Personally, I settled with colocation. I pay $60/mo + $2k one-off for the initial hardware + say $150/5y/4TB HDD, which, for 80TB of storage over 5y comes out to a total of ~$88/mo, or $0.001/GBmo. Even if I was to store 3 copies of everything instead of doing erasure coding, my costs are roughly half of what Backblaze B2 charges. And my disks are fully online block storage, not necessarily object storage.

Speaking as someone who hasn't tried colocation, would you care to recommend your provider or explain who the Linode and the DO are in that space?

I'm sharing a cabinet with a few other people in one of Hurricane Electric's datacenters. A full cabinet is $400/mo and we split it 7 ways based on space and power consumption.

If you can't get together the people for a full cabinet (which I think is definitely worth it, the more you buy at once, the cheaper it gets), there are a number of providers that offer single server colocation. The only one I've really heard much about though is Joe's Datacenter, which is in Kansas City, Missouri.

Unfortunately there isn't really a "Linode" or "DO" in the colocation space anymore. There used to be https://prgmr.com, which offered transparent pricing based on power and bandwidth usage but they've stopped that I believe. The few global providers are generally geared towards multinational corporations or at least businesses more than small groups or hobbyists. You're really best off with the smaller regional providers near wherever you are.

There is no de-facto colocation provider. Peer1 is an example of one that has a lot of locations, but it's not by any means dominant.

Every major urban hub has a multitude of providers. Some offer expensive, but reliable service, others cheap but you get what you pay for, and there's a mix in between. Talk to people and ask for tours. Every town is completely different in terms of scene.

Assuming that your are not in the middle of nowhere, it is typically best to stick with a somewhat local or regional co-lo facility. Discovering who else has space there is typically a good indicator or quality. Find out where your local hospital or university has their boxes. Keep in mind this is your hardware and it will occasionally need a new HD or fan. You do not want to fly across the country for that. Also their are a wide range of security options for co-los, so weigh those options.

IwStack (by Prometeus) can be customized quite a bit storage-wise:

http://iwstack.com/price-calculator.php

Delimiter has a clever solution for $10 a month. They call it Slot Hosting. You ship them a disk and they'll host it and attach it as additional storage on a VPS. https://www.delimiter.com/slot-hosting/

I use it to host an 8tb drive that I then attach to a dedicated server I rent for $20 a month.

$30 a month and I have a server that I use for Plex (16 cores, 3.2ghz, 32 gigs of memory, and 10tb of total storage).

Thanks for sharing - this is really cool.

Which hosting plan are you using for Plex?

[Disclosure - I work at Delimiter]

Thanks for the recommendation, slot hosting is a good solution if you need to get a disk online with a VPS attached to it.

We've had a lot of people asking how you did this so just to explain:

Slot hosting $120/year ($10/month) - that gives you one 3.5" or 2.5" slot to send your disk in. You can aggregate up multiple slots and aggregate the VPS resources into one large VM. This is great for things like ZFS and software RAID. The VM is KVM based. https://www.delimiter.com/slot-hosting/

The server is this: Dual E5420, 16GB RAM, 1TB or 2x500GB HDD $200/year ($16.66/month). https://cc.delimiter.com/?cmd=cart&action=add&id=1718

Slot hosting uses dedicated (non-contended) CPU/RAM/Disk resources so you can run Plex directly on it if you have enough slots aggregated to give you the RAM.

Why annual pricing? I'd do it if you had monthly pricing.....Since the payment is annual what happens if 2 or 3 months in I decide to change and need something else?

Slot hosting moved to annual only as administering a customer-owned equipment adds an additional layer of headache.

Its not cost effective to have customers who want to load up on 'warez' cancelling their service, shipping back their disk to get it back the next week empty for another round.

We're positioning this as a long-term storage product where customers don't want to be paying the hosting company each month for disks.

Can I buy slots and use them from your Cloud Resource Pools, (e.g. for $10/mo on top of the $6/mo for the Pool)? That's an interesting configuration...

The slots are just a VM + Disk. You can setup the VM to serve the disk contents by NFS or CIFS etc and mount it on any of our products.

Holy, why do you need that computing power for streaming shows/movies?

Transcoding... x265 CPU encodes on even a 2hr video can take several hours on a very fast intel CPU. I've been playing around with nvenc vbr 2-pass which has been giving me good results, not too much larger than the cpu encode in minutes.

But if you want something hosted/online, then it gets harder.

No way to offload it to the GPU? I remember reading about ffmpeg extensions for doing that...

I mentioned using nvenc (nvidia encoder), but didn't mention that is was using gpu... cpu encoding is much slower.

Depends on how many concurrent streams he's doing and if he's running anything else.

$30 a month and I have a server that I use for Plex (16 cores, 3.2ghz, 32 gigs of memory, and 10tb of total storage).

That sounds amazing. Where can I find this?

It works out at $26.66/month but its based on two annual products:

Dual E5420, 16GB, 1TB HDD (or 2 x 500GB): https://cc.delimiter.com/?cmd=cart&action=add&id=1718

Slot hosting: https://cc.delimiter.com/cart/slot-hosting/

Thanks for the response, but the spec is half of what you listed. you linked me to a 2.5ghz, 8core, 16GB machine. No obvious private network options, so pricing is pretty much middle of the road. Although the slot option is cool.

There is no private option with Slot hosting. I think the OP is confusing 8 cores + 8 threads as 16 cores.

I am in the same boat. Need additional storage for less random-access demanding data, like photos/videos. If you check the Linode forums, they have begun to work on block storage option, like Vulr/DO. But it is still early in the design phase and they don't have an estimate launch date yet.

Slightly off-topic, but EC2 doesn't really scale independently if you compare it to GCE. We let you combine 24 vcpus with 39 GB of RAM, 3 partitions of Local SSD and a few GPUs, all independently (though the ratio of RAM to vcpu is currently bounded between .9 and 6.5).

One of our easiest pricing wins is against say the i2.* series ("Were you trying to buy more flash or more cores?") and now against their p2.* GPU series. It's pretty liberating to actually mix and match.

Disclosure: I work on Google Cloud.

I used to use Linode for some projects, and really appreciated their speed and server quality, which seemed better than Digital Ocean at the time.

But after using them for 12-18 months, and losing several days of data due to the 2015 DDoS, and reading about more and more security issues, I switched back to DO and haven't looked back. The performance differences aren't noticeable to me, and I'd rather have my hosting through a company with a better security record than Linode.

We moved >1K$ hosting per month from them to StormOnDemand because of the lack of transparency on security issues, particularly the PagerDuty incident.

This was heartbreaking because Linode's value (performance, reliability, support, price) is almost impossible to match if your requirements fit their VM configurations.

But in the end, we did not want to risk another security or DDoS fiasco. We estimated that the risk was high that they would be targeted again, we could not believe their promises to get better at face value considering previous transparency issues, and we did not want to tell our customers that a company that had experienced security issues for the past three years had suffered from another attack[1].

They seem to have invested quite a lot in their networking infrastructure (kudos!), but I believe they still use their old coldfusion applications.

[1] Non-technical customers often ask us why we are not hosting on Amazon because they heard it's where serious companies host their servers (!). We used to explain why Linode was a more cost-effective choice, but Linode was not a nice name to google in early 2016.

>but I believe they still use their old coldfusion applications.

For now, yeah. But the new Manager mentioned in the blog post is just a client for our new API, which is written in Python.[1]

You can even check out the new Manager now if you want: https://github.com/linode/manager

It's still in alpha (and separate from our regular service), but we're hoping to have it out in beta soon!

[1] https://engineering.linode.com/2016/04/12/Announcing-APIv4.h...

DO isn't any better, just hasn't been targeted by any serious attackers yet.

until very recently, they didn't allow using a custom kernel (except via kexec hackaround) and were quite slow updating their kernel for security patches. they repeatedly gave random dates for implementation, then repeatedly pushed them back, then eventually just ignored users on this issue for years.

their images were also poorly sanitized, leading to the well-known problem of SSH host key duplication, which was the case for years.

> just hasn't been targeted by any serious attackers yet.

Source? Just because they haven't announced any successful security breaches, doesn't mean they haven't been seriously targeted.

In fact, considering the amount of times my random dedicated server instance (not hosted at DO) gets hit with random attacks, I'm sure a large provider like DO has had numerous serious, targeted attacks against their network/servers/control panel/etc.

> Source? Just because they haven't announced any successful security breaches, doesn't mean they haven't been seriously targeted.

True, and in fact the only basis for my statement is that they have historically taken security so not-seriously that it would be surprising if they were in fact able to withstand advanced attacks, given that even the most secure organizations are often unable to do so. (see: every talk at Black Hat)

> In fact, considering the amount of times my random dedicated server instance (not hosted at DO) gets hit with random attacks, I'm sure a large provider like DO has had numerous serious, targeted attacks against their network/servers/control panel/etc.

this statement is just as baseless as mine. perhaps even moreso, since the two numbers seem to have nothing to do with each other. one could just as well say "my server gets lots of bogus SSH attempts, so banks get robbed a lot".

>DO isn't any better, just hasn't been targeted by any serious attackers yet.

As the official authority on serious attackers I can confirm that his is in fact not true.

In addition to that, they even had security issues where people could use testdisk or any other file recovery tools to recover files which would often belong to another customer.

http://venturebeat.com/2013/12/30/iaas-provider-digitalocean...

PSA: If you're using the Linode API, the 2048 plan is now PlanID 2 and the new 1024 plan is now PlanID 1. The other PlanIDs were incremented by one as well. I wish they hadn't done that, but there it is.

That's really not cool.

That sucks. Those PlanIDs should never change.

Well, they've changed in the past... it's fine as long as the plans corresponding to the PlanIDs never get smaller or more expensive. This particular change violates that assumption and so is definitely not fine.

The hardcoded plan (PlanID 2) that I had in my program was pointing to a 4 GB 2 CPU plan (4 GB), now it's pointing to a 2 GB 1 CPU plan (2 GB).

There is an API method to query plans: https://www.linode.com/api/utility/avail.linodeplans

Presumably you can select the relevant ID based on plan properties (cores, RAM, etc.) rather than hardcoding IDs. Though I completely agree that once a plan has a specific ID, as long as that plan is available the ID should be constant.

Yeah, that's... not practical.

What if they upgrade the 1024 plan to 1536? (That has actually happened, by the way.) Now you're extracting the number and comparing it to a range?

> Give me the plan where the RAM is >= 1024 and < 2048.

What if they upgrade the 1536 plan to 2048?

> Nevermind, just give me the plan with two cores.

Okay, do you want the $20/mo plan or the $120/mo plan?

> Fine, sort the plans by cores and memory and give me the smallest plan.

Oops, they introduced a new plan with one core and less (i.e. not enough) storage.

I'm not saying it's impossible. Just impractical.

It's not as difficult as you're suggesting. Figure out your minimum specs and get the cheapest plan that meets them.

Or, the API's plan IDs could correspond to specific plans, and never change. New plans mean new IDs. Like how IDs are supposed to work.

The plans themselves have changed many times before. They become obsolete over time. What if your plan ID becomes unavailble, or the specs change in a way your application can't handle? Don't shrug off your bad design decisions on someone else.

Deprecated plan ID requests should return two plan ids (new_price <= old_price and new_specs >= old_specs). Then a bit of retry logic means the code will still function sanely until the hard coded ID can be updated (or it may continue to function indefinitely).

> What if your plan ID becomes unavailble

The API should throw an error, so I can adjust my scripts, instead of silently provisioning a server I never intended to provision.

There should be a deprecation process when a plan is discontinued, so this doesn't happen unexpectedly.

If AWS can't provision me a t1.micro, they don't just go and give me a m3.small instead.

Why have there been no Diep.io updates lately? Would you like to join our Diep.io Discord: https://discord.gg/vJVqhTf#discordbutton?

Yep just got bit by this as well. We're heavy users of the API and this was a surprise. Would have preferred some notification about the change OR better yet, just create these instances with new plan ID's.

That's just... ugh. Who thought ordinal numbers were a good idea for plan ids / instance types? Seems part of what you are paying extra for with AWS/GCP etc is competent engineering.

The obvious alternative would be to make the plan IDs reflect the resource limit, e.g. 1024, 2048, 4096, etc.

But that doesn't work with Linode plans because they keep increasing the resources every now and then. They doubled the RAM on all the low-end plans a couple of years ago. Before that, they doubled the storage.

Another option would be to use random IDs that point to a specific bundle of resources. But then the plans change, and the previous bundle is no longer available for purchase.

It's hard to use consistent identifiers when the resources they point to change so often. Linode customers absolutely love those free upgrades.

I have just created an account to say this: CHANGING THE ID OF THE PLAN IS INSANE AND CLEARLY A BAD-DEVELOPER CHOICE.

I bet there is a piece of code in Linode codebase that has the plan resources mapped in a table, with a list of ID attached to the various sizes. The genius developer decided that it makes sense to have the "growing number ID" match the "growth in resources" assigned to the plan. In other words, the bigger ID should always mean MORE resources.

This leads them to alter IDs if they have to introduce plan-sizes that are small.

Imagine if AWS did this: you would see riots in the streets!

I would love to add a couple of gigs of RAM to an existing linode for a small monthly fee. But there's this dropdown on the "extras" page, "90 MB additional ram - $5.00 / mo", down to "360 MB additional ram - $20.00 / mo". Yes, those are MEGAbytes. As in, you can get a 4.36G/ram node for twice the price as a 4G/ram node. Very strange.

I've always felt that those options were mostly there as a stopgap while you migrate to a larger instance or figure out another plan. I don't think they want most users adding features a la carte as that would throw off their numbers and make it difficult to efficiently provision instances.

Those are legacy options and to the best of my understanding are not available on new accounts or after switching legacy accounts to hourly billing.

I recently switched from DigitalOcean to Linode and could easily shave off 20% bill. The migration of VM was much easier than I expected (use rsync - https://lowendbox.com/blog/how-to-migrate-a-hosted-server-in...). My only complain is, I would like to take multiple manual backups even if it costed little extra.

VM migration between the cloud hosting services is a big pain. I hope someone comes up with an app to do that.

Nice to see Linode stepping up again and competing hard with DO.

Other than the Christmas 2015 DDOS, I have never had any significant issues with Linode. Especially in the Dallas DC.

I've been with Linode for a little over five years and a satisfied customer.

That being said - their policy for handling DMCA notices is abysmal. They will network filter out your Linode until you respond to the notice.

I can't speak to that really. Never had it happen and my work's customers are in Commercial Real Estate so our services don't dabble with DMCA notices.

Hopefully they at least give you some warning before a filter is placed.

Digital Ocean really needs to cut their prices at this point. There's no compelling reason to pay them 50% more for the same thing.

Love seeing competition at work. Two times in a week, after the T-Mobile/Verizon rate plan wars :)

Rather than cut prices (I think $5 is low enough), I would prefer if they would beef up their specs. For $5, DO gives you only 512MB vs 1GB from Linode.

6-of-1, half a dozen of the other. Halving their prices on all current instances gives you the same specs at the same pricepoints.

I don't particularly care if they keep around a 512MB/$2.50 instance at that point.

I would, however, love that price point. It would be nice if I could run a few light-weight services on a VPS for $2.50 a month and then put my websites and other projects on a beefier VM.

I currently use Linode, but if DigitalOcean offered something at that pricepoint they would be getting $2.50 more from me a month than they currently do.

Pure guess, but I'd imagine that payment processor fees don't make it particularly profitable to offer price levels much lower than $5.

One thing that could help is having the ability to pre pay/load credits, so you could have a small discount that way

This is the same theory that NearlyFreeSpeech.net uses.

Linode / AWS already offer prepaid discounts.

Where Could i find that ? For Linode AND AWS Discount.

For AWS look at "reserved instances"

I'm not sure about Linode

For Linode: https://www.linode.com/docs/platform/prepaid-billing-and-pay...

Make it so you have to buy at least 2 months then

Or even pay up front for the year. Perfect for the hobbyist market. I'd probably buy four of them.

It's more about support costs if I were to guess. How many support tickets does it take for their profit on $25 a year to disappear? (if $2.50 a month is discounted to $25 a year).

I always like prgmr.com's stance on that. "Please note that these plans include $2.50 worth of support"

ARIN charges for IPv4 address allocation (https://www.arin.net/fees/fee_schedule.html) and I read that below $5 a month it becomes unprofitable for hosting companies just because of the cost of leasing IP addresses. If your services can run on shared hosting there are a bunch of hosts for less than $1 a month around.

Charge a dollar for ipv4, or ipv6 is free?

Well, you get a 10 Gbps pipe with DO, while only 75 Mbps with Linode, lol!

Appears to have been a typo. It now starts at 1000 Mbps on the smaller instances.

Linode is 40Gbps IN, 1Gbps out for the $5 instance. The 75Mbps was a misprint.

From Soh (Linode) in the comments [0]: "We’ve deliberated and brought the floor for all plans to have 1000Mbps. Thanks for your feedback!"

So it appears that 75Mbps was the initial speed but after some internal deliberation (possibly influenced by feedback), they raised it to 1000Mbps.

[0] https://blog.linode.com/2017/02/14/high-memory-instances-and...

Had not seen that, thanks!

We did an on-the-spot upgrade to that speed, the new floor for all plans will be 1000Mbps.

The blog should be updated now, might need a refresh to see:

https://blog.linode.com/2017/02/14/high-memory-instances-and...

That 10gbps is highly variable.

I use one of my DO instances as staging for file transfers. 2 days a month, I can barely get more than 30K/s from their sf data center to comcast. Other days, I get 5M/s. Their network is shitty.

And I don't think I'm being rate-limited; it's 1-2G transfer per day.

Only new locations (FRA, SFO2) have the 10 Gbps pipe, and I am constantly able to push 2-5 Gbps through it in Frankfurt.

Keep in mind that bandwidth bottlenecks can happen on both sides - depending on the peering that is chosen between Comcast and DigitalOcean, there might be a significant oversubscription between two peering providers that's causing this.

DigitalOcean might happily be able to push 10g to their next hop, but beyond that things are significantly more outside of their control. This is mistaken as 'throttling' a lot, but is actually just ISPs not investing in peering to handle their peak demand. Similar to the Netflix and Cogent issue that happened a couple of years back.

At times, I can get better speeds by tunneling or proxying using a DigitalOcean Droplet and downloading something from overseas than I can doing it directly. The path taken from my home internet to the DO Droplet uses different peering than my path to Europe, and the speeds are faster overall even though it creates more hops.

Why? DIgitalocean tends to be easier to use and more reliable. I'm ok with paying a little more given these factors...

Easier maybe, but my experience of their comparative reliability has been the complete opposite.

I don't have the real numbers in front of me, but some back-of-the-phone math says we'd save something like $640 a month for equivalent RAM, which is our big limiter.

DO is slicker, but it's not 8k/year slicker. I can do better things with that money.

I don't get the whole VPS thing anyway. 5$ for playing around is great but anything that costs more is just not worth it.

60$/month for 16GB, 1CPU, 20GB disk for VPS... in comparison a root server with real remote console access, 16GB, 4 core AMD, 4TB disks, unlimited traffic costs me 30EUR/month (30$/month).

Or for 60EUR/month (60$) I could get a root server with 64GB Ram, 2x500GB SSD, i7-6700, 30TB traffic.

What company?

Mostly Strato and Hetzner.

I've pointed this out to people on numerous occasions, and the push-back I always get is concerns about timely provisioning if they ever need more capacity and/or new servers in case of catastrophe.

OVH et al can get you servers quickly if they have them "in stock", but supply fluctuates wildly, especially on the cheaper end.

Just being able to programmatically provision a server at a moment's notice—even in normal day-to-day (i.e. non-catastrophic) situations—is a critical tool in my toolbelt.

You can take a snapshot of a VM, you can move it to a different host, you can spin up a new instance to test a major upgrade, you can clone a cluster for a few hours of testing, you can scale up and down...

You pay for the scalability. Your 4 core AMD is always a 4 core AMD unless you migrate to a new server. Your 1 core Linode can become a 20 core Linode in a few minutes, for just a few hours while you need the capacity, and then go back to being a cheap 1 core Linode.

And one server will only ever be one server. For many deployments it makes more sense to have several smaller nodes distributed between different geographical areas than one big node in one datacenter. And/or multiple nodes in a a single datacenter so you can perform rolling upgrades.

What is your definition of a "root server" and where do you rent them? Do you mean a dedicated server? I've checked a few providers and none of them come close the costs/benefits you're listing. OVH's EG-16 (4c/8t, 16G, 2x4T) is $79/mo. 1-and-1's L4i (4c/4t, 12G, 2x1T) is $80/mo.

SoYouStart (OVH's lower tier brand) has 4c/4t 32 GB 3x120 GB SSD) for $49/mo. Or same price, same ram, 2x2TB and 8t with a slightly slower clock.

That's interesting. Thank you.

I realize that I am doing that thing where you try to rationalize your preconceived notions, but here are some high-level observations:

* Looks like there are normally setup fees associated with provisioning new servers, although they are suspended at the moment.

* Support seems next to non-existant, which I suppose is not surprising considering they're a low-cost provider.

* SSD-based servers appear to be frequently out-of-stock.

* If you need a KVM attached to your server, it is $30 for 24 hours or $200+ for a week.

* 250Mbps bandwidth (presumably in and out) cf. Linode which is 40G in and 1-10G out.

All that being said, my interest is piqued. I could see this being a good fit if 1. you have a more-or-less dedicated sysadmin, 2. need a lot of storage and/or memory and have solid sizing requirements ahead of time, and 3. cloud (VPS or otherwise) isn't an option due to cost or other facts. It could be great for running your own VM or container cluster. Thanks again for sharing!

Yes a dedicated server. Mostly german servers from Hetzner and Strato.

https://www.hetzner.de/us/hosting/produkte_rootserver/ex51ss...

Depending on how the host configures things, VMs can be much more reliable. If I know my data is striped onto a 16-disc array, and can be live-migrated onto a different storage box in case the RAID controller goes funny without me having to care or even notice, that's more than I can reasonably set up with a $60/month server. Couple that to live migration of the VM process itself if the hardware it's running on goes wonky, and you've got a recipe for as much uptime as you want.

All this does depend on the VM host actually implementing these things properly, but they are out there.

I love seeing cloud providers offering better deals, the more the merrier! However, I still have yet to find a better deal than the one Scaleway offers. ~$3/mo for a dual-core, 2GB RAM, 50GB SSD host. They also support terraform, rancher, swarm, etc. Been using them for a few months and haven't had a single complaint.

I probably wouldn't host a blog there. Their ToS[1] has, among other things:

- A prohibition against the propagation of data, images or sounds that may constitute defamation, an insult, denigration, or an infringement of privacy, image rights, good morals, or public order

- Users are required to use decent and respectful language

- Users are reminded that they must update software without excessive delay when a security failing is noted by the user or the software publisher or Scaleway

This reads like: If you might offend anyone by the standards of the French, or might use strong language, or don't run the latest version of everything, this isn't the service for you.

On the positive side, they lack the litany of restricted services that most hosting companies I've seen provide, and I don't see any prohibitions against using your paid for resources to their maximum, so that's good.

US customers should note that all of their prices are in Euro, so -- for exchange rate shenanigans and currency exchange fees on most credit cards. There's no selector to show prices in USD.

[1]: https://www.scaleway.com/terms/

On the other hand, being hosted in the EU means the data is (at least in theory) out of reach for US government agencies.

I was concerned about this as well. I use Scaleway but I don't host any websites with notable content there.

I haven't heard much about these terms being held against users, so I'm not sure how much action there would really be behind this type of stuff. With that said, it is definitely more comfortable to be on a host that doesn't have terms that are this strict.

Aren't they slow Atom C2750 cores though?

I believe the cheaper ones are the even slower C2550s. Another worry is the flaw thats been affecting the whole C range:

https://www.theregister.co.uk/2017/02/06/cisco_intel_decline...

They have also X86 alternatives since 2016 (and a third offering since a few weeks)

Atom is x86. I wasn't referring to their ARM servers.

It would be really fascinating if you added GPU instances with GTX 1080 cards. These cards could allow you to make the prices much lower than those of AWS GPU instances that use K-80 and make it a perfect fit for Deep Learning applications that don't require double precision.

OVH.com is selling 2GB VMs for 2.99 Euros per month. No idea about over-provisioning or any comparison with Linode though.

I'll say this as someone who's pretty happy being on Digital Ocean - Linode's support is amazing. If they had had a $5 price point for one-and-done projects forever ago, I would have never used DO.

OVH basically tell you not to use these for production applications. I have 3 that I use for as gitlab-runners and as a dokku development server.

I was planning to use one for a side project. Can you give a link where they say that?

I have one for a few months and according to htop, the VM is up and running for over 90 days.

I've been using one for a side project for a while and they're perfectly fine if you're ok with one core and 2gb of memory.

Sure, below is a link to some unhappy users posts. If you look on LowEndTalk, people commonly discuss how sub-10MB/s disk IO on the VPS SSD 1 plan is common, and many use these as cheap Tor relays, as OVH does not care what users do.

I'd rather something like ImpactVPS or Linode, as with either I know things will be decently managed, unlike OVH. Don't get me wrong, those VPSes make great dev boxes or Tor relays, they just have no performance or reliability guarantees AT ALL! Do not risk your business on them, spend a few minutes to find either a smaller provider with good management (generally at the same price point or lower), or pay up for Linode or AWS.

Link: http://www.webhostingtalk.com/showthread.php?t=1470503

The $3 VMs are different from the cloud instances. You get local storage with the $3 VMs which work great, while the cloud instances have some weird Ceph cluster that is making problems all the time. So the $3 VMs are way better than the regular, expensive cloud instances.

I use OVH. Their support is horrendous though. For example, I put in a support request at mid-day this Saturday to ask if their geo-ip feature where you can select which country you want an additional IP to appear to be from, meant that the IP address actually terminated in that country. They didn't get back to me until this morning (3 days later), and the response was a bit vague.

I bought some new IPs 4 hours ago for my VPS. I've had an automated response to say they've taken payment. Do I have the IPs? Not according to the control panel. I imagine I'll get them at some point in the next 24 hours. shrug If not, I'll put in a support request to ask where they are and maybe wait another 2 or 3 days for a response.

When I signed up a few months back, I purchased my first VPS on a Saturday. At no point did they tell me that my order needed to be manually checked. I eventually got my VPS on Monday morning.

Does anyone even work there on the weekend?

Amateurs

Meh, doesn't sound too bad to me. Its not like your machine has failed and they are not responding. You use OVH/Hetzner with the understanding that the great prices come from minimal support and the hardware will be used or desktop class. Hetzner support is strictly standard office hours. If you want instant responses, you can pay 10x at another provider but you chose not to.

That is only partly true. First of all OVH has different offerings (and I expect that I can get a new cloud server any time?) and Hetzner replied instantly to us when we had a server down on a Saturday. You can even call them by phone. Hetzner is no cloud and mp fluff and feels a bit aged, but the support is not bad and for normal questions within hours on working days. PS: I do not work for them ;)

The thing is, it wouldn't have cost them anything for them to have sent me an automated email to tell me that my order needed to be manually checked, and I would get it within X hours/days. In fact, because they didn't give me this information up front, I ended up logging a support ticket to find out what was going on.

It wouldn't have cost them anything to tell me up front that IPs aren't automatically provisioned and required manual intervention and it would take ~24 hours before I get them either. Luckily for them, this time I didn't log a support ticket, because I just assumed that they would have crap processes and that's why it was taking so long. I mean, why did it take so long? Where they manually checking my order again, even though I used the same payment details? Does somebody need to manually pick some IP addresses from a spreadsheet? Why is this process not entirely automatic?

This is the sort of thing that should be automated for low cost hosting systems, precisely to prevent having to provide unnecessary support.

Had an OVH account. They suspected server was hacked. Nuked my account and hence the websites on it (it wasn't; many were affected at the time and it was a false alarm). Moved all the content to a folder (that was on a server/path to which they just gave me FTP login) and I had to get that data by some period of time. Tried to get in touch with them at least 8-9 times. Every time received a copy-pasted response on how to download my data. Just that. Called them once and I was asked to contact on email.

That was it. They also own Kimsufi. I had one account there too. Cancelled both of them and moved to VPS providers and it's been 3 years w/o complaints (my usages were/are very small time though).

When I checked around on hosting forums, looking for ways to contact them, everyone used to laugh as in "Support and OVH? Lol".

Is this a bold move by Linode, or am I the only Linode customer (7 years going) who chose them in part because they steered clear of the bottom of the market?

I think it makes good business sense. It comes less than a year after they bumped the $10/mo instance to 2 GB RAM. They must have realized they were losing market share just by dint of there being no $5/mo option to compete with DO. Even though this is the bottom of the market, as you say, some of those "one-and-done" projects must hang around and need to scale eventually, and it's clearly easier to scale up on an existing platform than migrate to another. I know this takes away one of the two or three big reasons I ever have for using DO instead of Linode!

I'm sorry, but this pricing doesn't put them anywhere near the bottom of the market. As someone that has been buying from the VPS market for almost twenty years now, Linode cannot in any reasonable verbiage be lumped in with the riffraff crap vendors at the bottom. I could rattle off a dozen or more companies run by thugs and thieves that truly define "bottom of the market" for VPS, and Linode, DO, Ramnode aren't even on the same planet.

I know that I switched to Ramnode when I had problems with another VPS provider back in 2013 or so when one of the Admin panels was hacked across a large number of VPS providers. Ramnode was down for a few hours, the other VPS was down for a few days.

Some feedback: I've been using DO droplets since mid 2013, reliability is excellent, no unexpected reboots at all.

With the help of Ksplice the Ubuntu Server droplet has achieved 555 and 401 days uptime without downtime (could hang on a bit longer but later decided to reboot once every 3 months to address security concerns).

DO support has been responsive and friendly, DO keeps (slowly) delivering new features such as private network, Load Balacer etc. For existing $5/m DO users, I don't think it's worth the hassle to migrate to Linode (or Amazon Lightsail), the performance difference will be unnoticeable for most people's use cases (personal web hosting, strongSwan based IPsec VPN, etc.).

A good reference: https://joshtronic.com/2016/12/01/ten-dollar-showdown-linode...

Will provide the feedback to DO and see if they can match the Linode offer (I am sure they will do something).

I've been using Linode for many years, long before DO was around. Very happy with Linode. I tried DO, but I didn't see them making a compelling enough reason for me to switch over. Linode has a strong track record.

I enjoyed that link to the $10 showdown. Linode's VPSs usually preform quite well. It did take Linode a while to get into the SSD game, but when they did, they got the IO right.

Appreciate the link, I have a $5 review up as well now :)

Since my posts are straight benchmarks, I don't touch on the support side of things typically (mostly because I only have had support interaction with the companies I use).

That said, I've had exceptional support from both Linode and DigitalOcean over the years. Always responsive and friendly.

Interestingly enough though, yesterday my Linode went down due to a power outage in the Atlanta data center. I'm a huge fan of Linode, but something like a power outage at this stage in the game seems a bit like amateur hour.

Scaleway recently added new cloud servers. Example: 10 cores, 60GB RAM, 700GB SSD, 1Gbit/s Unmetered = 90€

Maybe one day that ColdFusion atrocity of a control panel will be hurled into the sun and something better will take its place. Until then, yay, cheaper high-memory instances.

Redis appreciates it.

ColdFusion or not - Linode's hosting panel is one of the simplest and flexible management interface I've experienced. Digital Ocean's is too simple, and some other providers are too infuriating. Linode is that porridge that's just right.

> ...one of the simplest and flexible management interface I've experienced...

Well, I'm really questioning what you've used then. It's atrocious. It's right up there with GoDaddy in terms of dashboards that are needlessly obnoxious.

Digital Ocean's may be simple, but there's nothing wrong with that. It works. It's clear what it can and can't do. It's not cluttered up with confusion.

For example, on Linode you cannot delete a Linode instance anywhere but the main view. You must go back to the main listing, carefully look for the one you want to remove, then click the remove link and double-check you clicked the right link. If you have a lot of instances and you cycle them over frequently enough this is a real hassle.

Likewise, there's many occasions where you get kicked back to the index page for no reason. There's just so many unresolved little things that, over time, grate on you considerably. It works but it could be considerably better. It has not evolved much since launch, that's very concerning.

Digital Ocean's interface, to use one example, has evolved considerably. Amazon's AWS dashboard may be a monstrosity but it's also becoming better and better organized over time. Linode needs to remember that their dashboard is important and invest in it.

Maybe all you ever eat is porridge and you're okay with that. Fine. Other people demand some real food now and then.

We do think our dashboard is important, and we are investing in it. As a matter of fact, you can even watch the development take place yourself because the new manager is an open-source app. :)

https://github.com/linode/manager

That's really good to hear. Please give us an option to use it as soon as it's available for production accounts.

Wow, $5 finally. Must have gotten tired of DO and Vultr taking those customers

Might finally get me to give up my ChunkHost account for throw-away servers.

Why not 2GB Ram and 2 cores, easier for CPU demanding on a $10 plan?

Thanks to amazon lightsail we now have $5 linodes :-)

Lightsail $5 plan has inferior specs to the Linode $5 plan.

Yup $5 shoot out between Linode vs DigitalOcean vs Vultr https://community.centminmod.com/threads/10437/ :)

75 Mbps?? Are we in 2005? :-(

They just upped all plans to a min of 1000Mbps

I think that was a misprint. The floor on all plans are now 1000Mbits. :)

Can we stop ignoring the fact that Linode has been hacked several times and responded terribly in every case?

That is why I stopped using Linode. It wasn't the fact they got hacked, as much as they minimized and lied about the extent of it.

Same here. Why aren't we more unforgiving about such lies? It is, after all, a commoditized service.

In the current environment, I personally believe that it's safe to say that everybody has been hacked to some degree. Linode being open about it, even if not having the perfect response, strangely makes me feel better. Being a customer while this occurred and not feeling any impact makes me feel even better still.

Having to go in and change my password is hardly an exceptional event on today's internet.

They were not open about it. When their ColdFusion system got hacked they didn't tell anyone until weeks after the fact.

When you search for "linode hack," Google suggests "linode hacked again." They were hacked in 2012. They were hacked in 2013. They were hacked in 2014. They were hacked in 2016. They will be hacked in 2017.

> they didn't tell anyone until weeks after the fact.

Unfortunately, this is fairly standard practice in the industry. Companies want to make sure the vulnerability is closed, positively identify what was compromised, who was affected, what legal liability exists, and so forth.

Weeks is, frankly, pretty quick to go through that process.

> There's not much more to say -- they are dead.

Huh. Funny, I'm still hosting things there; their prices are competitive, there are no rumors of acquisition or shutdown... Seems quite alive to me.

> They will be hacked in 2017.

And, as I stated originally, I have no reason to think they will be unique in this.

I don't think there's any reasonable arguments left to defend Linode with.

https://news.ycombinator.com/item?id=10845278

There's a clear, undeniable pattern of incompetence here.

I also suggest reading this glassdoor review: https://imgur.com/sJd56AT

And this thread: https://news.ycombinator.com/item?id=11136743

>> this is fairly standard practice in the industry. Companies want to make sure the vulnerability is closed, positively identify what was compromised, who was affected, what legal liability exists, and so forth.

This is so far beyond malicious and incompetent it should be illegal. What you really mean is "Are we obligated to report this to our customers, or can we cover this up and get away with not letting anyone know we've (maybe/probably) been hacked?".

Customers' entire businesses are on the line. As in, a company can literally go bankrupt and/or be forced to shut down if the hack affects them. The only acceptable resolution is to warn customers within ONE HOUR of knowing that their company MAY be at risk of a hack. One day is already too late. A week later means that no pre-emptive mitigation was even possible, and it's simply too late to even try and protect oneself.

IT IS STRAIGHT UP NOT ACCEPTABLE, TO NOT IMMEDIATELY INFORM A CLIENT OF A __POTENTIAL__ THREAT TO THEIR BUSINESS. ___POTENTIAL___, NOT ___CONFIRMED___.

Companies like Linode are so busy trying to cover their own PR asses, that they don't understand just HOW CRITICAL it is that their clients be instantly informed of any potential threat. They think their business's reputation is important, without having a single clue that their entire business's success relies on their clients' businesses being safe. Informing all their customers that there is a 0.00001% chance that their account has been compromised FAR OUTWEIGHS the eventuality that even a single account was in fact compromised.

They just don't get it. They are prioritizing their own business's PR over their clients' businesses' well-being. And so a single hack reported weeks after the fact, without any early warning having been raised, completely destroys all credibility. A hosting provider should be put out of business after a single such failure to immediately warn clients of even a remote possibility of a problem.

tldr; Providers like Linode who prioritize confirming their liability, before so much as even considering issuing a warning to their clients that they may have been compromised, should not be allowed to do business. It really is as simple as that, and frankly anyone who continues to host with a provider that failed to raise any warnings until weeks after a potential hack deserves whatever business-destroying event happens to them next time. You cannot trust a company once they've purposely postponed releasing crucial details of an incident. Quite literally: in the future, when you find out you've been hacked on Linode a month after the fact... what the fuck did you expect? Precedent indicated this was the likely outcome... you got exactly what you stayed signed on to experience!

I'm really curious to what people plan on doing with such insane amounts of ram.

Maybe it's possible to scale with just RAM! Fire up a uWSGI + Flask and up the RAM as traffic increases.

database servers, redis and memcached I'm guessing. All of which are basically designed to take a memory-time tradeoff.

Yes, this is what I use my high-memory servers for. Giant memcached and redis slabs.

You can run a memory-based DBMS. That will easily eat all the RAM you can throw at it.

AH! Linode! You used to have a $5/mo plan and you got rid of it. I found and switched to DO. I would have stuck with you but I only use it to mess about and 10$ was too steep. If you make it $5 Canadian I'll switch back.

It's $5 USD, just like Digital Ocean. If Digital Ocean works for you there's no reason to switch back. They're equal cost now.

In the past they kept changing their pricing. I was paying for the smallest one (static site and a few prototypes) around $150/yr. Then they advertised that per-use would be mandatory and cheaper for everyone. I kept the site there barelly receiving any hit, using 0% of cpu and network. Endedup paying well over $250 after 12 monthly charges.

Unless I'm mistaken, per-use means you aren't billed for the server when it's powered down.

If the server is up, even if you use 0% CPU, you're getting charged.

Per-use is a lot cheaper for people that spin up servers to meet demand and kill them once things slow down.

From Linode's FAQ:

"If My Linode is Powered Off, Will I Be Billed? If your Linode is powered off, but is still added as a service on your account, you will still be billed for it. This is because Linode maintains your saved data and reserves your ability to use other resources like RAM, transfer, etc. even when your Linode is powered off. You will be billed for any other active Linode service, such as Longview Pro or an extra IP, as well."

[0] https://www.linode.com/docs/platform/billing-and-payments

"If I'm not in my hotel room will I still be charged? I was at a ball game for three hours! Also I wasn't in my apartment all weekend, can I get a partial refund?"

Seriously, people do not understand leasing and renting.

Question here is about they using the new model as a smoke curtain for the price hike.

All their marketing during the change was how it was cheaper. When it clearly was over double the price. And your argument is "it is cheaper if you are not using". So good luck going to a hotel and not booking any room so it can be cheaper.

It would never be cheaper for you because you want a single low-utilization server that's available 24/7.

It's only cheaper for companies that have lots and lots of servers, but most of them aren't provisioned most of the time.

I've used Linode since they opened and their pricing has been extremely consistent, it has never gone up or been randomly changed.

They introduced per-hour billing recently, that's the only fundamental change to their billing, and that was something you had to go out of your way to opt into if you wanted it for an existing account.

If you provisioned a $20/mo. machine then yes you will pay $240 per year for it. This is how math works.

Reply on news.ycombinator.com