All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Linode vs. Vultr
In 2019 I switched to Linode and cancelled all my low end boxes and simplified things with a single vps. I've been happy with it, but lately I've had routing issues in the evening from my home (Asia) using wireguard.
For a long time I used Fremont or Dallas, which was mostly ok, but no more. That leaves Singapore or Osaka. Tokyo has no availability, though I used it a year or so ago and had some issues. Osaka gets routed through Singapore, unfortunately. And Singapore has issues. By issues, I mean high latency, packet loss, and low speeds in the evening.
So I switched to Vultr, Tokyo, which seems to be mostly ok so far. Sometimes it has some issues in the evening, but restarting the wireguard tunnel a few times will get a good route.
All else being equal, I prefer Linode, but the location and routing are more important.
These are some of the things I've noticed that are better with Linode.
1) At least for me, they don't block port 25, whereas Vultr does. I guess Linode does it for new users now, too. I don't send much email, but I'm using Mailjet (free) right now for smtp.
2) Linode provides Slave DNS service, whereas Vultr doesn't. I'm using he.net for free slave DNS now.
3) Linode provides an extra /64 ipv6 block (or /56) on request, which I used for wireguard. Vultr only provides one /64 block, so I have to use ULA with wireguard. I guess you can get a portable ipv6 block for $3, but that doesn't make sense to me for a $5 server.
4) Vultr's 25gb disk space doesn't provide as much space as Linode's 25gb disk space. It's reported as 23G. I'm sure there's a logical explanation for it, but I'm not sure what it is. I'm close to the limit so it matters.
Also, Vultr's $5 server is using older generation servers, though I'm not sure it matters.
Comments
That makes quiet a sense to choose Linode over Vultr
That’s HighEndTalk
I honestly couldn't tell the pros vs cons between linode and vultr since my use case only care about bandwidth and latency most of the time.
back then when I started using Vultr simply just because they have terraform integration, this is really helpful for in-demand deployment just using few lines of codes.
https://github.com/opentofu/registry/tree/main/providers/v/vultr
and today i discovered apparently Linode also have it too, but I won't be trying them out as long as Vultr still works, simply because I'm lazy. and I didn't notice any huge gap between the hourly pricing anyway
Vultr:
and df -h says:
dmesg:
Linode:
I'm not sure why they appear that way. Both VMs I spun up are Debian 12...Vultr's image seems a lot fatter.
Interesting. I also spun up Debian 12 on Vultr and then restored from borgbackup at rsync.net. So there could be some extra stuff from the Vultr image. I know I had less free space after the migration.
Linode has massively improved routing ever since they joined Akamai since they now use the Akamai backbone.
In Japan they now have KDDI, Telstra, PCCW and NTT which basically includes all of the major asian backbone networks.
https://www.linode.com/speed-test/
What ISP are you with?
PLDT Fibr.
I would have tried Linode's Tokyo location again, but it's at capacity and unavailable. I did try Osaka, but the routing is not good.
Oh for SEA yeah that'll be rough. Have you tried their new Indonesia location (although I think it's more expensive?)
FWIW this is probably more the fault of PLDT than anything else, most SEA ISPs don't really peer well.
It could be the their ddos protection limiting you.
Linode's Jakarta location seems to go through Singapore as well, so it doesn't help.
You could try v2ray vpn with cdn (i'm using cloudflare) to overcome bad routing. Dunno about latency but the good thing is, you can change the cdn's ip that not throttled yet with your ISP.
I don't think anything is getting purposefully throttled. I think it's just congestion of certain popular routes during prime time evening hours.
I spun up a Linode in Osaka to have a look at the difference in disk space. I think the lower amount of free space in Vultr is due to the way the file system is setup. The Vultr file system has way more inodes.
Vultr:
Inode count: 6285088
Block count: 6422267
Block size: 4096
Inodes per group: 31904
Inode size: 256
Linode:
Inode count: 1605632
Block count: 6422528
Block size: 4096
Inodes per group: 8192
Inode size: 256
Routing through Linode Osaka is not good, though, 90ms latency vs. around 45ms to Tokyo, since it gets routed through Singapore.
Linode also provides DDoS Protection standard in all DC's. Vultr charges $10/month
How do you guys sleep at night, liable for bandwidth overages should something occur? With no way to programatically limit overdues at the provider’s end?
That’s the biggest issue I have with these providers. The base prices are okay.
And for that reason, I’m out. None of them.
Edit: If I remember correctly, in the past, you actually could set a hard limit at Vultr. Guess the feature didn’t help them (as it seems they removed the feature awhile back.)
How does Digital Ocean compare to both Linode and Vultr?
And get this:
According to the Vultr docs:
"For billing purposes, each month has 672 hours, and bandwidth allocations are made hourly throughout the month. An overage charge will be applied if you exceed your currently-allocated bandwidth at any point during the month."
So 2TB means you get about 3GB per hour (2048/672). If on the first hour of the month you download 4GB, you're liable for a charge because you exceeded the bandwidth credit rate.
WTF?
This is not strictly true because all Vultr accounts get 2TB free egress per month (available at the start of each month), pooled across all instances in all regions.
But yeah the policy they have there is I guess to prevent people from spinning up an instance for an day, push a bunch of traffic, then destroying the instance for a low cost.
The upside is that all bandwidth is pooled, so theoretically you can push 10+ TB per month on a single small instance if you have other instances accumulating bandwidth.
DigitalOcean offers a feature called bandwidth stacking, where they aggregate the total bandwidth for all your VMs. This way you can sleep well. UpCloud also had this feature.
@Dazzle no I couldn’t. With a 10 gbit port, a lot could be used fast. Especially since multiple VMs = multiple 10 gbit ports.
To be honest, I thought it would just stop if it hit the bandwidth limit. Ignorance is bliss. I never came close to using all the bandwidth, though. At most, 100-200GB per month.
I could write a script to monitor the bandwidth and close port 80/443 tcp if a limit is exceeded, or shutdown the instance. Most of my bandwidth is for personal vpn use.
I'm still not sure I understand Vultr's bandwidth policy, though. On the one hand, I can see it accumulating available bandwidth daily. But there's also a global allotment which includes 2TB. So do overages not occur until you exceed 2TB+plan allotment?
Vultr by huge margin.
Linode as we know is long time gone. Akamai took it, molested and made corporate cadavre.
How many droplets do you need for this to happen?
Really?
From what I've seen, they've added tons of locations, improved the bandwidth blend by a lot and, to top it all of, decreased bandwidth pricing significantly.
How did they make it so much worse?
34 VMs range from 1GB to 16GB. Additionally, there are some S3 Spaces, I believe they also count toward the pool.
Why are you on hourly providers with resource needs like that? Why not go dedicated?
How is their ddos protection?
It is for various clients, managed service.
DO is our lowest standard for these specific services after we migrating from AWS and Azure.
Selling DO is easier than any other newer brand, clients are likely conducting research before deciding to go with us. So, yes, it may be about brand value, etc.
I see, may I ask what's managed? Server management i.e. everything they might run or you do specific stuff?