New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
You can add a downtime notification for free if youd just setup uptimerobot :P
As for managed vps I recommend Liquidweb and their HeroicSupportTM
Get live chat help within 59 seconds, get phone answered within 59 seconds, helpdesk 30 min reply.
No managed needed, he can just setup redundancy, either within DO (not recommended by me, but by them) or within multiple DCs around the world and providers. This also "load-balances" things to a point, depending on how complex the setup you can do/afford.
AWS is the champion of uptime, indeed, but not even them are perfect.
DO could setup some affinity zones, but, you should be able to use different geographical zones to make sure you dont land on the same node...
Jesus, I'm wondering how a "Super Moderator" can say to "GTFO"...
Anyway, a side-note: my $2-$4/Month shared hosting providers DO send me notification when they reboot stuff, install new stuff which could lead to downtime or when they know the server is having issues like kernel panic whatever...
Thanks to all of you that actually DID try to answer my question instead of going completely nuts / crazy like the few couple ********** (apparently some need a bit more sleep and should stop drinking 2 liters coffee / hour)
I'm with the op on this, the hosts system should know which customers are on which node and they should certainly send out notifications if they know they have an issue with that node.
It doesn't matter if it's a $1 service or a $1000 service, typing an email saying node XXX has issues we're working on it takes all of 30 seconds and if they're using something like WHMCS they can just send based on the product group.
Getting such a notification can save a ton of wasted time as you then know it's an issue with the provider and don't have to sit there checking your app/system.
Edit:
I'd imagine it would save on some support tickets as well.
@benjamin74
If you want reliable hosting partner with guaranteed uptime, dedicated support person, bells and whistles try these providers:
Believe me, they are on the next level and you will feel that.
Once more I am recommendinh liquidweb
Great service!
Correct in terms of cost also!
If you say you've been fine with VULTR why not just use them? VULTR has always provided me with great uptime (I use their NJ location), support, and they're constantly rolling out new features.
Are you sure the downtimes are actually an issue on DO end of things? I have droplets in a couple different datacenters and it is very rare I have any type of downtime.
Jesus may read this forum, but He doesn't post.
Leave my son out of this mess!
This post is getting hilarious! Ahah!
Anyway, as @time4vps suggested, those companies will take care of you. Not within DO budget, tho.
You can also check Linode, I use them in production.
Please kindly leave me out of this mess. I am okay with @raindog308 cursing.
Debian. Thx. Love you all!
that is so weird. says joined like 10 minutes ago but doesn't appear in the activity.
He's Jesus, man - he does whatever a Jesus can.
It seems like Codero.com's Linux Cloud Hosting offers DO budget. And even more than that, they offer some more disk space with some plans, some more bandwidth with some plans and some plans are even cheaper than DO's. It seems their 8GB plan is the best offer for a buck.
I knew it!
must be exception for Jesus, he is 'Übermensch'
Would it be easier to create a phone app to handle those downtime notifications along side with email? If there isn't one already that is. Most people have those "smartassphones".
Guessing this is a question mostly for @jarland
You've his the nail on the head here on most key points relevant to the OP. Seems like a no brainer, really. Interestingly enough, it's a very rare breed of LEB provider who does that. Most likely because they simply get away with it and see no need to bother. As you've correctly pointed out, the cost, whether in terms of time or money, would be negligible.
BTW, nice to see someone here, who comprehends what he reads
As opposed to most others, who are going completely offtopic, recommending managed service, amongst other things, in answer to OP's valid criticism of lack of notification about planned, or at least anticipated, downtime in service which he is otherwise perfectly happy with.
To be clear, what the OP referenced was not planned or anticipated downtime. Though, still a valid criticism.
Now, most hosts wouldn't notify you twice of an issue that took 5 minutes to resolve. They would say "Hey we're rebooting" or "Hey we rebooted" but not nearly as likely say "Hey we're rebooting" and "Hey we rebooted."
The problem here arises when you've planned out for yourself an easy way to send a notification that says "Hey we rebooted" as the sole notification, and the situation is more complicated than that. That is, of course, a much more rare occasion but still one to plan for. Let's say a RAID controller dies and it takes 4 hours to get things back up, it's a terrible idea to say "We'll send out an email saying we rebooted in 4 hours."
So what we're talking about here is a lack of workflow in the system for saying "Hey guys, this node is down, and we're working on it." As it stands, the best method I have to do this is to open manually a ticket with each customer on the node (and don't think I won't). Obviously, this is a gap in the workflow. It's a clear need, and it's on my shortlist for improvements to follow up on.
I guess I'm the only one finding this thread a bit blasphemic.
LunaNode. Not in US, but close enough (Toronto, Montreal). Passes all other OPs requirements.
Auto-monitoring by HTTP/TCP/ping/DNS and alert by email/SMS/voice/HTTP hook. Min interval 60sec.
As for support responsiveness, only had one ticket several q&a in it, responses within 2-3 minutes.
They also provide managed support, quote: Server management: open a ticket below in "Managed Support". While our services are unmanaged, we provide 24/7 managed support at $60.00/hour (billed in 15-minute increments). ... We are also able to automatically intervene on your server if a service fails ...
We usually post in forum and in whmcs with progress, if only one node is affected, an email goes out as soon as we know what is going on with some basic information and link to forum for updates.
Mailing thousands of customers when there is a DC wide problem, though, that is not a good idea, first because it takes a long time to send those mails and we may have to send another before the first batch was finished, second because, by the time we find out what is going on, it is usually almost fixed.
An email, though, might not be the best solution, it usually requires checking your mail to find out. You can as well use uptime robot or nodeping if you want to receive an email when things go south.
I stand corrected
Nice to hear, but this would be overkill (read: not sustainable) at LEB prices. I was implying a well thought out automated solution…
Good point. Sorry, obviously haven't considered the numbers involved.
Except that the point here is not notification of downtime, but notification of the reason behind it (e.g. it being a DC issue, as opposed to something failing on the customer's box) and ETA fixing it. Understandably, not important from (especially momentarily distressed) provider's perspective, but a world of difference to an informed and competent customer with emergency plan in place.
Maybe an automated, one off email, then, with a heads up: "Houston, we have a problem! Please visit … or follow us on Twitter for updates" (except customers from many countries, like China, where Twitter is blocked) ?
Or even a status page somewhere, providing it's actually kept up2date and customers are made aware it exists(quite a few providers fail at this).
Edit:
I'd also suggest off network.... just in case it's a core router.etc that goes bang, and plan for some way to update it if your DC goes dark (E.g 4G/Backup internet access for the staff)
Unfortunately theres still a time and costs, which often arent negligible for LET providers charging only a few $ markup.
Time to keep system uptodate (often when time is tight as staff are investigating the issue)
Cost to develop (or deploy, test and maintain existing solution) said system.
That being said I quite like the "Droplet * rebooted due to physical node issue" messages, it might not cover all cases, but its quite a nice email.
The big providers could start integrating with all the new Android notifications apps like Pushbullet (based on GCM AFAIK)
Hm, I used to have a droplet with DO and it didn't go down once when I was with them.
Unlucky node, I guess?
I have downtime monitoring in the form of SMS messages - I receive a message that my website is down.
No problem, really - if you're using CF, just call their API and switch your root record when your monitor detects downtime.