[Mahn]

For what is worth, I've been a Linode customer for a few years and never really felt their management front-end was in urgent need of an update. It's solid and does the job for us. My only complains to be honest are 1) the lack of storage dedicated nodes, as OP points out, and 2) the price, it gets very expensive very fast as you scale and spin more nodes.

We are currently considering moving to dedicated machines but only because of 2), otherwise we would happily be their customer for life.

[threeseed]

> never really felt their management front-end was in urgent need of an update

Even when it was hacked multiple times and customer VPSs compromised e.g.

http://arstechnica.com/business/2012/03/bitcoins-worth-22800...

Nothing is worse than spending weeks securing every aspect of your VPS only to have incidents like this appear. And worst of all ? To this day Linode never clearly said what happened or what they did to prevent it happening again.

[mumphster]

Not sure why this is being downvoted but the whole Linode hack incident was a huge factor for me to switch all my VPS's over to digital ocean.

[brobinson]

I went to Vultr. DO's interface is just as bad if not worse than Linode's. :(

[lugg]

Its not that bad, it has been updated frequently though. Seems to be getting better.

I do remember it being quite lacking back in the day.

[brobinson]

I haven't used DO since 2013 and I probably won't since I'm extremely happy with Vultr, but it's good to know that it's been improved!

[prirun]

I chose vultr over DO because DO used so many external web services to run their site. Every new page at their site required me to turn on 5 more domains in NoScript. Vultr only required me to enable their site. The performance at vultr was also very good for what I needed: test servers.

[mulchkin]

Why Vlutr and not something like Scaleway (https://www.scaleway.com/)?

I've recently switched from DO to Scaleway and would never go back.

[brobinson]

Actually, I did try Scaleway when I saw them on HN like 9 months ago! I recompiled a Go program I was working on [1] for ARM and benched it against Linode.

It was literally 10x slower than Linode even after playing around with Go's concurrency level to find the fastest runtime, and even with the dataset in memory. :( ARM just wasn't the right arch for what I was doing.

I ended up going with Vultr because it has the $5 pricepoint for hosting tiny websites and tons of datacenter locations. Their CPU performance and network speed were great in my tests.

[1] https://github.com/robinsonstrategy/go_backtesting_simulator...

[astrodust]

It "does the job" but it's not as pleasant to use as it could be. A lot of the navigation is pointless and confusing, information is often buried several levels deep, and there's missing information on some of the detail screens you need to go back to the main listing to find: Instance type is only shown on the main list, not the individual instance tab, for example.

If GitHub had never improved their site since they launched it would be awful. Every time they move things forward I'm happier to be a paying customer. With Linode I reluctantly use them, but for new projects I'm using other services that work better.

[cookiecaper]

Same here, moving to a dedicated server from a Linode this week. For slightly under 2x the monthly cost I'm paying at Linode, I can get a dedicated server with 8x the memory, infinity more transfer on the same size pipe (unmetered 100Mbps connection), 20x more disk space, and 2x the CPU, and all dedicated so no worries about phantom performance problems related to other tenants. Lish &c are a big plus for virtualized nodes, but I think I'll live without it.