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
He'll get it fixed
@unused, we are the on the same page
Problem is the commits and the combined costs. Cross connects although cheap, still add up in the big picture.
Do I think BuyVM could afford these places? Unsure. Know it would eat into horse food.
We tend to think of BuyVM as a huge operation. They aren't. I mean they have customers and gear but it's probably combined, three racks (see their status page for idea of the server pool size). It is peanuts compared to many of the more common hosting household names. They run a damn efficient operation though and are some of the nicest folks I've dealt with in all my years (other folks mileage may vary).
I can't think of one budget VPS provider selling services out of Coresite, Equinix or Telx.
" a company with hundreds of millions+ behind them is not only going to make sure the infrastructure is right from the start but can do what's needed to maintain it, and respond in ways to issues that a smaller player can't."
This isn't a truth. If we apply that logic to business in general we wouldn't have bailed out the biggest banks and financial companies in the US and the biggest auto company.
All facilities experience outages. I wonder which facility has the true best uptime (no power failures, no bandwidth failures, etc.).
The only glaring differences may be 1. Staff (having enough seasoned staff on site) 2. Over built redundancy and contingency planning to work around problems. These two pieces are downright expensive. The people are very hard to find and mock failure survival training / games are hard to simulate.
You are spot on about the one big thing a place like Telx or Coresite has that others don't and that's their private peering exchange. But those are a love/hate feature depending on your needs and who you ask. I've seen plenty of totally retarded routing in these solutions, latency that was laughable, and just a complete I don't get it.
These top notch facilities make sense when you JUST HAVE TO BE THERE. There might be 10ms away from Wall Street for trading or similarly close to Washington, DC, master control. That's what those places mainly function to provide for (big finance, trading, government).
Looks like Vegas is back online.
This is the one area we'll have to agree to disagree - my point isn't about the size of the company, but how much cash they can put into infrastructure. Where a smaller (but good) facility may go for lower rated equipment, or less capacity, less onsite spare parts, staff/training/maintenance contracts/response time etc. and an even lower tier/budget facility will go for used/refurbished/repair of parts etc. a company that has the funds mostly likely won't skimp and has an advantage over a smaller facility.
That said, I don't think the network is a big issue- (ref my original comment)- what's potentially disheartening is the power. Hopefully the guys just oversubscribed their circuit and tripped some breakers/it was a minor issue.
From what I recall though FH has a history of power problems (I could be wrong, I'm recalling WHT posts from a year or two ago.) that would be annoying. Even so, at LEB prices, power dropping out a few times a year may not be a big issue. Just look at how many people stick with HE in fremont-- like Linode
corexchange dropped power 3 times in a year. Those on single power went out, including Incero that's popular here.
There is normal equipment, pro equipment then exotic. Any facility is going to have pro equipment. Perhaps some facilities have exotic, but rare. Any hardware will only function normally as good as the humans maintaining and configing it. Some things like electric can be built with more redundancy and idiot fusing to avoid massive arc welding of humans to chasis. But it tends to be if it is in a data center generally acceptable.
The spare parts is often a sore spot for facilities. Applies to everything from the server pieces to the HVAC. Most places aren't stocking a full datacenter of spares, but have a good 24.7 rolodex of local vendors who have the pieces in the same metro.
Staff is number one or are zeros when problems happen.
As far as budget places using used and refurbs... Bet it happens. It is fine in some situations. But I don't think it is when it is core power stuff, batteries, gen sets, etc.
Now power is a common fail point. There isn't any good way to test power or simulate failures. Power gets tripped and batteries fail or genset doesn't come online fast enough, well, lots of issues. Real facilities with bigger customers probably are mandating and/or customers are dual fed power to each rack. That should eliminate most of the power problems we see as feeds to facilities tend to go down just one at a time typically unless a huge grid level brownout or many distribution network failure.
What happened today at Fiberhub, from what I read, was a 200AMP breaker installed on the wrong side of the generator. Unsure who was playing electrician and if it was just an honest mistake or what. Again this stuff happens even at the best facilities. B leg power to racks of those effected would have made the outage a non issue.
HE Fremont, yuck. No thanks. They've had their share of outages power and otherwise.
Good example @concerto49. It is a good facility too. But shamefully near quarterly power drops suck.
Lots of new providers have no clue about power redundancy and assume it's built into the facility as every facility has all this power equipment and SLA claims of total uptime.
B legged power to me is a silly hack although it is wise to do. Facilities need to truly build the power redundancy in upstream.
Cost to customer for the B power can get pricey real quick. Have the cost of the second feed monthly from the facility, plus whatever power controllers or dual power supplies in all your gear.
They already been booted from LV or something and heading back? That why all the downtime ? But this down time is horrible
They've been up for the last 2 hours.
Networking is still messed up on mine, but I'm assuming they're still fixing things.
Same
Pinging .. with 32 bytes of data:
Request timed out.
Request timed out.
Request timed out.
Request timed out.
Ping statistics for 209.141.39.159:
Packets: Sent = 4, Received = 0, Lost = 4 (100% loss),
after experiencing few times up & down today, my vps seems ok in last 2 hours, hope things getting better soon,
Other than the fact that the DDoS protected IP on my VPS isn't working, it seems to be online. Hopefully that should be up soon.
ProviderService has a private cage with Equinix in Dusseldorf. ProviderService has LEB priced VPSs itself, plus GetKVM has servers with ProviderService. So there are plenty of low end boxes in Equinix at least.
Of course you can argue that anything in the Old World doesn't count because free peering etc, but whatever man.
addons now working
@heiska, good post. The old world counts.
Checking out GetKVM.
Does Internap Texas count?
darn my forum was down for days. lost quite a lot of viewers. luckily I found a backup. already moved into another cool provider.
@aolee you didn't plan ahead for the -scheduled- expected downtime/move ?
NY seems to be completely down.
BuyVM went up, so we moved SolusVM back onto it... only to have it go down again straight after ;(
(we're on KVM LV 08)
@Nick
yeah i can confirm that.
regarding my ping stats my vps is unreachable since 12:42 CET (GMT+1)
I was in FMT2 for ~2 years. Only ever had 1 power outage.
So does anyone know when the DDoS protected IPs will come up and why there down as I think that is the main issue for most at the moment?
http://twitter.com/aldryic/
i cant reach irc. not even with the ip i could use before. pls pm me if you have a working solution.. thanks
It was moved to the NY location for the move. NY appears to be down at the moment.
@Nick
I thought they have it running on both locations