All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Ethernetservers - last minute DC migration
It seems like Ethernetservers decided to preform last minute DC migration on Sunday.
Emergency Migrations - All USA VPS Customers notification mail arrived few minutes before the migration started. Just great...
You are receiving this email because you have one or more active VPS' with Ethernet Servers.
As many of you will have noticed, there have been server stability issues in recent weeks following a data center facility fire.
To protect the integrity of our company, and our customers, we are conducting emergency server migrations.
This is not the way we would like to be handling the matter, as in an ideal world, we would provide considerable notice upfront, however, we believe it is in everyone's best interest to act sooner rather than later.
With regret, this does mean that IP space is changing.
Unfortunately, it is not feasible to provide individual status updates upon start and finish of the migrations, so we will be updating https://www.ethernetservers.com/clients/serverstatus.php with our progress.
To prioritize this task, we ask that all non-urgent support requests are deferred until this is complete.
Our sincere apologies for the trouble caused here.
Kind Regards,
Ethernet Servers
My server was migrated some hour ago, but no IPs swapped, so it's as good as down. I hoped that this will be done automatically with the migration.
Anyone experienced the same or that's a glitch? I don't want to open support ticket right now when they are most likely overwhelmed with all the work.
Comments
As it stands right now, there is no "clean" way to do these migrations, it's all manual.
We attempted to upgrade to SolusVM 2 first to allow for automatic migrations & IP swaps, but ran into a number of issues, of which SolusVM's development team are investigating and informed us they will have more information tomorrow.
I can only give my apologies for this situation. It's far from how I want to be doing business.
Mine is under migration for an hour now.
@EthernetServers, my servers have IPv6 addresses allocated to them.
Are these also been reallocated?
Solus 2 has broken IPv6 the last time I tried.
Almost every provider using Solus 2 is getting on the list.
We encourage customers to leave providers who chose to use crappy panels or hypervisors.
VirtFusion is panel of the year.
Right this moment it does not look good at all.
From the unplanned last-minute migration on Sunday to the nonworking IPs after the migration and extensive outage which still last. I hope they won't screw up also IPv6 part once stuff comes online.
where is he migrating you guys to?
They're in a mix of Equinix and Digital Realty facilities.
any info on the actual upstreams?
@EthernetServers is not production worthy
It's easy to sling mud, can you be more specific in your claims? From what I understand this is a result of dedipath issues and they're taking a little upfront pain for a lot of long term stability.
I agree, EthernetServers should have a teleportation system to quickly teleport all of their servers to a different data center with 0 downtime. I am losing $5 billion per minute that my servers are down.
I agree, the issue is caused by Dedipath and if a provider is proactively working a long term solution, they should be appreciated, nobody thought that Dedipath would do this ... even until now no explanation of what's going on with them ..
I mean dedipath is kinda shit, I definitely thought they were going to end up imploding eventually. Using low quality upstreams is risky in general.
I don't see the benefit of slinging mud. People clearly used them for a reason.
People used them because they were the cheapest way to get bandwidth from INAP, not because they were good or competent. I'm not slinging mud, just being objective.
This migration isn't handled well no matter how you look.
There was zero, yes zero notification (mail came some hour before all the stuff went down) upfront, on Sunday and day later all our stuff is still down.
It's not about losing millions it's about migration, not some unscheduled outage.
How hard is to warn people at least some day upfront and prepare yourself to do things as smoothly as possible instead of doing it impulsively and completely unprepared on several locations the way they did it with all the crap day later still down. It's not like dedipath kicked them out right this second and plan it properly wasn't possible.
As far as I understand people are questioning whether dedipath is going down under. Given their recent outages you might risk them going down again and perhaps never coming up again. Then the real outrahw would begin.
Could it be handled better? Sure, but given the circumstances I understand that he wants to migrate ASAP and not give some warning where there is a window for the servers to go down again.
And taking all the stuff down immediately for unspecified long-time solve issues how?
From where I stand this outage ain't better than dedipath outages.
They saw a tunnel ahead, but with no light at the end of it, so they made a strategic decision to take a different route.
I'm not their client, but support their decision to move according to the given circumstances.
That's not an issue and no one objects to that. The problem is that they did it in a "VirMach" manner and that sucks. It's migration for fuck's sake, not some unscheduled outage.
Don't just take all your stuff on several locations down one day and then see what will happen.
Current situation:
We're still working with support at SolusVM to address some bugs that are preventing us from importing our SolusVM 1 installation into SolusVM 2.
Most customers on the few servers we've migrated manually have new IPs assigned now, but some remain. Working to complete this as quickly as possible.
The plan is to finish the remainder using automation available in SolusVM 2, which will significantly speed up the time it takes for the remaining servers.
As we all know ethernetservers is a one man show, any future downtime is just going to cause him loss of income due to clients cancelling, he can keep his cost low due to the past reliability that his hosting offers, he will rather risk taking more losses in this migration rather than risk anymore downtime that is beyond his control.
I did not realise they used dedipath. Withdrawn....but then again always go directly to DC
Though I appreciate the step they are taking.
But man, re-doing setup due to IP change on multiple VPS with them will be painful.
Going directly with a DC can sometimes be very expensive and generally you need to get your own transit from somewhere, which can also be very expensive. Depending on the DC, it can sometimes only be worth it if you have a few racks.
Generally it makes sense if you have that sort of scale and have a few racks.
There are some DC's like Hetzner FSN/NBG/HEL that directly offer some smaller colocation packages and IP transit for customers with a smaller scale, but that's generally rare.
Down for 17 hours so far.
I am also wondering about that.
I have been with hostus.us from 2015,but now all vps cancelled at the moment,as always long time no response from them for tickets.
For ethernetservers,also start to use from 2015 till now,still have 5 vps with them for long time plan,at least,they care about their customers.
Very much so. I know some people aren't happy, and I'm really not happy either, believe me. This is the last thing I want to be putting customers through.
Just like hostdare,it was good before,but after it is owned by another company,it becomes bad,one time,it was offline for long,even the hostdare webiste offline,no one knows what happen and no explain on it.
For such case,if you really care about your customer,you even need to make a simple homepage to let your customers know what is going on.So i also leave hostdare,no longer use.