All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
RamNode 4 hours downtime yesterday evening
I've been a huge fan of RamNode for a while and host many important servers with them. Last night, they had a failed switch in their ATL datacenter and my servers were down for over 4 hours. I couldn't even log into the control panel.
One of my servers has an uptime of 350+ days and this is the first time it went down for any amount of time, but it's still pretty troubling to have a very important server go down for over 4 hours.
I guess it's a lesson learned and a good time to figure out how to setup a failover. I obviously have full backups but nothing setup to automatically switch over in the case of hardware failure.
Is there not real way they can monitor for these kinds of failing hardware before it happens? And is it even normal to take 4 hours to fix a switch?
Comments
Yeah... I was affected by this as well and was annoyed because I had set aside time to work only to find I could not.
This is really my first and only significant outage with RamNode and it was with a $15/year plan, so while my evening was ruined, I don't think it's a big deal for me in the grand scheme of things.
Stuff happens...
If this server was important, why hadn't you set up a disaster recovery plan ahead of time? Was it simple laziness? Or perhaps a bizarre assumption that a good host could never experience significant downtime?
they can pet the hardware raid card and ask it if it's feeling well?
no some stuff don't really give you a heads up of imminent failure. Also did you try to contact them about your question?
They have answered any kind of questions I have thrown to them.
And on one year uptime one good hardware failure (if that is what happened - I have no idea just from what you mentioned) I don't think it warrants making a thread here.
Some things you cant predict, 4 hours including trouble shooting and switch config restoration is fast.
If you want a host with redundancy at a switch level... pay more.
I'd sign right now for 4 hours of downtime per year at every ISP I have services with, if that was possible.
Nice reminder that computers are still computers, networks are still networks. No matter the size of your brand name or the amount of uptime you boast, every host has an area of weakness that can take your content offline. No matter how much redundancy they boast. Always have a plan B
verry nice
and about 4 hours yes is preatty good time. Depend of the issue
As soon as the downtime hit 2 minutes I would have been on here so fast to make a scam thread.
IT HAPPENS!
My 15/y big drive plan always go offline for <10mins daily around 2am. It's a backup drive so yea I am still happy with it.
Oh no RAMNODE had some downtime. PLEASE RUN FOR THE HILLS! /sarcasm
No refugee offers yet ?
never mind ,they are great.
@stab just out of curiosity what do/did you think was an acceptable amount of time to replace a failed switch?
4 hours of downtime, they are going to be deadpooled soon.
1ms. /s
The downtime should have been less than what it ended up being. As mentioned on Twitter, we are going through our setup to make sure future events have less impact. No, there was no way to know this switch would simply stop working.
4 hour per year is about 99.95% uptime. I doubt you can ask more than that for the money you paid. Honestly I didn't even notice the downtime on my backup server with Ramnode.
Waiting for "PreDowntime Sysops" that stops hardware failure before they act, reducing the downtime to zero.
time to put them in the eternal damnation category.
Did they every give an update on what caused the downtime. Must have been something big as they are pretty good at getting things taken care of.
Surprised this was even discussion worthy. Seems like something that would have just been brought up, and then let go of.
"Hey, was RamNode down yesterday for four hours?"
"Yeah, I think."
further discussion about TacoHost's opening
It's worth having it 'on the record'. If it happens more often, I think the track record of the host lends itself to how scathing the thread will be.
It's already 'on the record' on their status Twitter account: https://twitter.com/NodeStatus
neutral venues preferable.
If its not ramnode, most people here will curse it. But thank god its ramnode, 4 hours i dont feel it at all.
Context is really important here. RamNode is not a problem provider. Even being an "okay" provider really puts you in the doghouse in this situation. I suppose you can say it is the privilege they get for their ongoing reputation. Now downtime is downtime, and they don't get a pass for it, but they certainly don't have a history.
That's exactly why you need reputation..
Of course:
@Jack - We said early on that a switch failed and had to be replaced.