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.
UGVPS abandoning NC - forced migrations to GA with nearly no warning.
Just got a rather cryptic email from UGVPS that I think means my vps is getting moved to GA from NC and will be assigned a new IP... tomorrow.
Not much heads up there, UGVPS - less than 24 hours.
Anybody else got any more details as to why UGVPS is pulling out of NC? I'm still waiting for a response.
Comments
Share the email!
The email:
https://gist.github.com/NodePing/5122379
So UG VPS just became...yet another Colocrossing host...
Given the price that colocrossing is offering, I think it will be just a matter of time before all VPS providers in US to be colocrossing-based, LOL
I've been pretty happy with the service and performance I've received from UGVPS. The vps I have in NC has been one of the more stable from LEB.
Big change out of the blue kind of makes me feel a bit sideswiped though.
lol, she's acting in this mail like you're getting some gift or something. Not a word of apology for IPs change, location change, potential downtime, wasted time and all other hustle no one like. Just a "little" change, no big deal.
It's impressive how some hosts treat their clients as stupid.
I guess they figured out the RTO plan from Dacentec does not really worth the money, since those dual L5420s are quite a few years old and the network out of Dacentec is getting quite unstable recently as well. I have my test node hosted with Dacentec and received quite a few downtime emails recently (although all very brief, well within 10 minutes)
Can't say I blame her. Network has been getting better at Dacentec, but I've pulled it from my order form. I'm not leaving, but I can't put it next to Dallas like it's just a location choice.
If that's pingdom, I'd check twice. I've been getting false reports of outages in NC from Pingdom. Granted Ryan has had more trouble from his server there, but our VPS node has had extremely impressive uptime in both power and network.
Or just packing her bags like in UK case before...
It will most likely happen again.
Like I said, I've only had a couple of network downtimes with the vps there in NC in the last few months, and they were very brief. I don't use pingdom... <grin>. Kinda partial when it comes to monitoring.
We like to keep our hosts geographically spread and we already have a couple in GA. I'll likely cancel my service at UGVPS and look for an alternate east cost location - maybe something back in NC. Any recommendations?
I don't want to step on UGVPS toes here or look like I'm trying to lure you away, but if you're moving already and you want to stay in NC I've got space that I'm just not advertising. Node is performing like it's empty.
Getting rid of the aging L5420's is a definite positive but the network at ColoCrossing in Atlanta (at least with CVPS) has been the most unstable of any of my VPS's over the past 2 weeks. Frequent timeouts and packet loss, monitor.us is showing the VPS was unreachable almost 12% of the time during the past 2 weeks (see the comments in this thread http://www.lowendtalk.com/discussion/comment/221779#Comment_221779). They were supposed to be getting a new switch today and adding new capacity to alleviate the network problems...hopefully it does fix the problems in Atlanta.
Possible, I will place a uptimerobot check there as well in this case, thanks for the tip.
actually I am wondering why...I mean, if she is thinking of cutting the cost, I would guess she would have been looking around for a while by now rather than have sign the lease 2 days before her Dacentec server bills are due...
Nevermind I got that backwards. One of those days. Uptime robot was giving me the false reports. And it wasn't reports, it was one report, yesterday. My memory is getting worse every year.
Only making a quick post here since CVPS was mentioned, I will remain respectful to Crystal & UGVPS so apologies for my quick post here, just wanted to address this as it applies for the whole CC Atlanta network.
We apologize for the inconvenience that CVPS Atlanta customers may have witnessed the past few days, there was a bandwidth issue with CC in Atlanta because they were awaiting a switch replacement to come in.
The switch replacement arrived earlier this afternoon and after they replaced it today, everything has been very stable and on 1Gbit speeds.
Thanks,
Kevin
Seems like BW issues in all locations!
Please elaborate? The issue was only in Atlanta, but that was already all remedied today and I can honestly say all of our locations are now solid and running smoothly.
Regards,
Kevin
Yeah, I'm a NodePing fan in terms of fewer false positives. But man I wish you guys had a public status page ready...
And a better API! Why can't there be one call to get the status of all checks instead of having to call each check to get the status of each one. One API call verus 50+ in my case.
I hope not. Not advertising, but not something I'd do. It mainly happens for hosts doing 2GB for $7 type VPS unfortunately. The margins are very low and the risk is too high.
It's sad to see the trend.
NodePing,
I did send this email out this evening to all NC clients. I am starting migrations tomorrow, but that doesn't mean your VPS will be moved by that time. It may be a matter of a few days, i am starting with NCNode1 first then migrating NCNode2.
Given the price that colocrossing is offering, I think it will be just a matter of time before all VPS providers in US to be colocrossing-based, LOL
Colocrossing has a good US network but not entirely true, I still offer services in Coventry, UK and do not have any plans to discontinue that. I'm just waiting for more hardware to be built before I can offer more stock there.
This is most definitely not a cost problem, really the opposite as I am paying more for Atlanta =P I think anyone here knows that L5420's are old and you can get them pretty cheap on an RTO. I am migrating these customers to Atlanta servers that are on E3's with SSD-Cached Hardware RAID10. This is a huge improvement in hardware and obviously costs more to build. I am paying more money for my new Atlanta servers than I was with Dacentec. The main decision behind the move was because of the network, and also blocking mails on our ACL switch by default, and although it was lifted after i made a ticket, a week later mailing is blocked again and upon an abuse complaint they would block mailing again.
I just don't think they are very friendly for VPS hosts anymore. At the same time I want UGVPS to use updated hardware. After this migration is complete all servers will be on up to date hardware.
-Crystal
I can see where they're coming from. You can't be running a node with software RAID 10, the disk I/O was pretty weak.
Huh? Actually it has only improved drastically lately, from what I've seen.
They have had about 3 scheduled downtimes lately and that was due to network maintenance, where they were adding more network capacity or improving things.
I personally seen huge route improvements and speed increases. Went from 75 ms ping from Minnesota, to 48 ms.
We're working on that public status page.
There is one call to get a full list of current 'down' checks.
But it does mean it will be moved to another IP before Monday - which means I'm working the weekend to build this box out on another provider in NC before it goes dark. A few working days notice isn't too much to ask, is it?
Hostigation. I've a nagios box with Tim and it run well as a monitoring box.
My Pingdom trial expires soon
Hurry 
@UGVPS COV is out of IPs, atleast for the next 3 - 4 Months.
I don't really like UGVPS, was hosting my radio over there at their UK location, but my radio suffers from packet loss there every day, then it goes down for a few minutes... (or someone is abusing the network on the node)
Moved my radio over to @BlueVM today though
Yeah, this was a problem previously. It was like a domino effect on all three locations, because they're all connecting to the same routers, on the same line.
Coventry got more capacity added recently, and also on a seperate line to maidenhead. So heavy traffic towards maidenhead will no longer affect coventry and vice versa. Ref: (http://easevps.eu/announcements.php?id=7)
+1
Tim's services are a awesome; great network, excellent performance, straight-forward no-bullshit support.
Highly recommended.
Yep, CC Atlanta has been stable since the new switch went in yesterday.