Hacker News Reader

Comment by jordan0day | Hacker News Reader

> etc and/or folks who think they need to smash the (shared) resources out of server to get their moneys worth.

Wouldn't this be more likely on a higher-priced VPS (like Linode) than a lower-priced VPS (like DigitalOcean)?

I need to introduce you to a site called http://lowendbox.com - where scriptkiddies, 12 year old "web hosts" (like, a web host run by a 12 year old) and just weird (sorry, but I just don't get the personality type) people seek out to spend as little money on VPS's - often debating the merits of switching from one to another in order to pay $3.50/m instead of $4.

They then try to eeek out every bit of performance they can from the accounts they "collect". It's a curious sub-culture.

Heh, that's not entirely untrue - the majority of people on lowendbox and lowendtalk seem to have some sort of strange fetish for running things on as little resources as possible. I've seen some people there that actually use the LEBs for things like cheaper redundant hosting, etc., but they're the exception.

I've found they're very useful for running scripts I wouldn't normally want to have an impact on my own production VPS. A few times I've had people asking to run a trial of a new webapp (forum system, blog, cms), and these work perfectly for the task.

That's a presomptuous (and unnecessarily aggressive) you are making here. I certainly don't want to pay more than 3 euros a month for the use I am making of my VPS (it's still very useful, don't be mistaken), and LowEndBox was incredibly useful in finding a good provider for that price range.

People certainly have different uses than you (and your expensive 100's of hosts), that doesn't make all of them script kiddies and 12 year old hosts.

No, definitely not. If your top priority is performance per dollar, then no rational person would choose Linode. As a result, Linode attracts people whose top priorities are reliable support and good uptime. People who are worried about uptime don't run their boxes at 110% all the time.

> good uptime

Which is funny because Linode has pretty bad uptime.

When I was on their Fremont DC it went down all the time.

Apparently Fremont is their oldest DC and seems to attract the lion's share of the downtime. I've had 8 linode's running in their London DC for almost 3 years now and can't remember any DC related downtime. The only hiccup I've ever experienced is a host machine having disk issues once which meant they needed to move one of my vms. From being alerted of the problem to having the vm back up and running took ~45 mins with no intervention on my part.

I still subscribe to the RSS feed so no you're wrong. It's not exclusive to Fremont.

And nice that you were alerted of the problem. I often found out hours AFTER their issue was resolved what was happening.

tbh I have no idea as the London DC has been rock solid for me. On the odd occasion I've learned of Linode DCs having problems (usually here on HN) it seems to frequently be Fermont that's mentioned.

As for alerts, maybe they only automatically create tickets and alert users when there's a problem with a vm or the host it's on. Agree with you that doing the same for DC issues would be a good idea.

Fremont is a terrible DC for linode and it sucks if you are not aware of that when you buy a vps from them but I use the Dallas DC and haven't had any major problems to speak of in quite a long time (at least 5 AWS outages or more).

Fremont is by far the outlier here, its easily their most error prone DC. I'd highly recommend you move to another DC, all of which are much more stable.

Disclaimer: Customer for ~4 years, VPS in all DCs but Tokyo.

My node in London DC has 2 years uptime.

Reply on news.ycombinator.com