Linode Staff explanations and HE

I agree - I have been a member of Linode for over 2 years now and the HE network problems over the past month have caused me all sorts of customer problems.

The worst part is the non-communication from Linode - when we had an outage of about 18 hours in January due to some routing issue at HE with our ip's I couldn't get a response from Mike after an initial "we've notified HE of the problem". Even once the problem was resolved I got nothing and our support ticket was simply closed.

Yesterdays outage took us out for well over an hour all up - almost two. I'm concerned that the outages on HE are getting more and more frequent. For a tier 1 provider - I can't imagine how one (1) router playing up interrupts our network connectivity - doesn't tier 1 indicate redundancy of some sort….?

I've got to admit - I'm close to giving Linode the boot - you've been great up until now but things aren't looking so great at the moment and I can't afford to have this affect my business. It would have been nice to get some replies from you ;-)

JJE

While I agree on most points, I do have to remind you:

"You get what you pay for"

just what's that supposed to mean? u pay for bad router connectivity? give me a break! i dig self-managed systems, not HE crap.

@maven:

just what's that supposed to mean? u pay for bad router connectivity? give me a break! i dig self-managed systems, not HE crap.

Get back to me after you run a large DC w/ 0 problems.

that's very typical of someone who has done all that and been everywhere.

just to let u know that i don't care about your impeccable professionalism or track record which of course I'm in nooooo position to judge but it doesn't detract at all from the points in this post at all. ie. that you don't have a bad router connectivity that adds up an hour or two in a typical day and expect paying customers to be as understanding as your most accomplished self.

enough about members talking amongst themselves? what does linode think? hmm?

I have to agree.. overlord's posts here haven't been overly helpful in this topic which was intended more for linode staff.

I've worked for telcos and do appreciate the problems a tier 1 can have from time to time, but it still is no excuse to have unscheduled outages for hours at a time. Most likely outages are due to misconfigured routing that just cannot converge, in which case some techie needs to have his head hit with a BGP tutorial.

Problems at datacenters do happen, that's a fact of life. Of course, we'll never know exactly what occured over at HE, but I have contacted my sales rep there and asked what they're doing to prevent things like that from ever occuring again.

We were on top of the problem, as much as we could have been. Both Mike and I were in contact with HE throughout the night, each time we were told "it'll be fixed soon" and sure enough it was, although it would repeat some time later. The fact that the problems were intermittent and the promise that they'll "be fixed real soon now" at that time didn't seem like it warranted a forum post, but in retrospect we should have.

Next time, I won't be so hesitant to post a status message. My apologies, and thanks for your comments.

Regards,

-Chris

@monarch:

I have to agree.. overlord's posts here haven't been overly helpful in this topic which was intended more for linode staff.

I've worked for telcos and do appreciate the problems a tier 1 can have from time to time, but it still is no excuse to have unscheduled outages for hours at a time. Most likely outages are due to misconfigured routing that just cannot converge, in which case some techie needs to have his head hit with a BGP tutorial.

That's nice. You're entitled to your opinion. I just hate whiney, penny pinching who want HA for $30 a month and then complain when there's any slight hiccup.

For what I'm paying them, I think I'm getting a hellovalot better then you'd find just about anywhere else.

well i think it's nice too that you are entitled to living with mediocrity just as we feel we don't need to

While you can criticise Linode for not letting us know what was going on, it looks like HE weren't being very informative, so Linode didn't have much to tell us.

However, as kiomava points out here, the problem resulted in 90 minutes total downtime over the last few days, spread across a dozen or so bursts lasting between one and 30 minutes. Describing this as mediocrity seems a little strong.

@caker:

Next time, I won't be so hesitant to post a status message.

Any chance of getting those messages a RSS-feed?

I'd rather get too many instead of too few of those messages, but would also prefer it if I could limit "my" feed to just those events related to the server that I'm on.

Thats not a bad idea, and myabe on the same note have a host side email, ie sends email to everyone on that host, and maybe have alternative emails for people incase they are running their email on that host, and therfor can get it from another place..

cheers

NF

@tony:

Any chance of getting those messages a RSS-feed?
http://www.linode.com/forums/topics_anywhere.php

-Chris

@caker:

@tony:

Any chance of getting those messages a RSS-feed?
http://www.linode.com/forums/topics_anywhere.php

Great, just one suggestion…

There are lots of people out there using RSS (or at least can use it directly in their webbrowser), but, which like me, don't bother coming into the "community" before they get upset about something not working… and once here that RSS-feature isn't obvious for people not used to phpBB; so it might not be a bad idea to put together an official RSS-feed and alert new users about it (on the page where you submit a new support ticket might also be a good idea).

Yes, perhaps with a different feed for each host, so we can get alerts that are specific to us.

@Xan:

Yes, perhaps with a different feed for each host, so we can get alerts that are specific to us.

I really like this idea.

And there could also be feeds for issues affecting entire data centers (ie Hurricane Electric and The Planet). Or those items could just be injected into the appropriate host feeds.

I've actually had some outage stuff in the works that would display information for current outages affecting your host when you login as well as on the page where you open a support ticket. Having RSS capability wouldn't be a bad idea. I will mention it to Chris.

Any chance of an email alert to advise any possible problem on my linode. I could point the email to my sms gateway and then get paged about any problems :)