Hxxx
Active Member
Yeah I noticed, congrats.Sweet, we got a vote.
LET Q1 2014 Top Provider Poll Results
- Thread starter drmike
- Start date
DomainBop
Dormant VPSB Pathogen
I've been testing it for about a week and haven't encountered any problems yet but one week isn't really enough time to know whether the problems earlier versions had have been fixed...I like to test for a couple of months before putting something in production.
switsys
Active Member
Being 'verbal' as a customer doesn't necessarily mean anything.
I've been a member of LET for years, but I have only five or six posts.
Yet I have/have had accounts with around 60 different providers.
drmike
100% Tier-1 Gogent
Well, in order to see profile details one needs a LET account. Since they like to power ban mine, I refuse to play along and refuse to blow up their viewership numbers when I reappear.
Feel free to toss me a pre-setup account and I'll run through the profiles.
I never did IRC and god knows how impopular my posts are (I get attacked left and right not as bad as here, though).
I do believe the poll is a community , not an industry one, therefore people hanging around have services from providers popular there and are voting for providers popular there. That is absolutely normal, here BuyVM would win hands down any day and night, but it is worth noting that none of the CC brands managed top 3 in LET. At least since CC took over.
Last edited by a moderator:
drmike
100% Tier-1 Gogent
The poll is a ton of shilling. Not run of the mill vote for me petitioning either. Seen folks in the past asking their IRC homies for votes, heard of freebies for votes, I know cheating is going on.
To my point, anyone with large enough votes cast and average vote value of 2.0 or above smells of vote abnormalities.
Not pointing fingers at anyone, but some brands have clear cheerleaders who show up in 5 minutes of anything anywhere. Their brands aren't that big to justify the pattern.
A poll like LET's just isn't big enough participation wise to say much for real. It's whoever is fashionable in the community. But even calling it a community poll is really lying. Why? Cause LET and LEB have all these massive viewer counts, right? Like 100's of thousands of people a month, at least, right? But yet when it comes time to vote,
There were 557 votes cast for Q12014 by 244 different voters.
Do the math: 100k viewers a month (really low estimate) compared to 244 voters...
244 / 100,000 = 0.00244.
That 2/100ths of a percent of viewers deciding who is the "best" VPS provider in current 3 month quarterly.
Even if we slash the unique viewers a month to 50k:
244 / 50,000 = 0.00488
That's 4/100ths of a percent.
If we slash uniques to 25k a month:
244/25,000 = 0.00976 --- almost to 1%
Unsure how a poll would work around here. It would be equally as skewed/weird though since here most folks are providers (similar high density over on LET).
Doubt BuyVM would be the top. Definitely would have a strong showing. But so would RamNode, Prometeus and DigitalOcean.
Freebies always win the hearts of the LE people. Vultr gave free 2x credit and people stocked up on it.
Is the product worth it? I don't know, i've not tried it.
As I said somewhere else, Mr. Clarke gave a 50% discount on IP's, services, etc, and suddenly he had a multi page sales thread on LET even though he was banned and is on record trying to exploit multiple providers,
including us.
Well, that's how it goes. If you push a lot of discounts (be it constant or otherwise), have sales posts on a reasonable basis, etc, you can pull a lot of votes. This and the last few votes have had people pleading for votes on twitter, mass emails, etc, so it skews things.
I'd love to see us top 3 again and it wouldn't be hard. Ask some people on twitter and I'd end up with 200 - 300 votes easily enough. I'd then be driving traffic to a community that I don't give a single fuck about, though.
It's also easy to win with nice DD's. The DD's are gamed, though, due to tunings in cards. I won't go into this much since that's for everyone else to debug, but I ran into a lot of it in recent weeks.
Francisco
Last edited by a moderator:
drmike
100% Tier-1 Gogent
^--- I am 100% on what Fran said. The SSD DD games... haha I call what is going on FRAUD. Some folks have setups where DD speeds exceed the spec/capability of the underlying SSDs. Figure that out.
Also 100% on Damian's experience about being forgotten since not hanging out over there playing with the kids on LET.
Also 100% on Damian's experience about being forgotten since not hanging out over there playing with the kids on LET.
Not sure I consider that fraud, just people who have HW RAID cache.
Not fraud, but purchasing certain parts because of specific features to tune DD's
Anyway, this is fine. There's people that buy simply on the fact of certain benchmarks.
They had a bad past with providers due to reason X, so they want to make sure X is never
a concern. That's fine.
This is going off topic so I'll end this now
Francisco
Oh well, give the people what they want I guess. Yeah we have gotten off topic haven't we lolNot fraud, but purchasing certain parts because of specific features to tune DD's
Anyway, this is fine. There's people that buy simply on the fact of certain benchmarks.
They had a bad past with providers due to reason X, so they want to make sure X is never
a concern. That's fine.
This is going off topic so I'll end this now
Francisco
drmike
100% Tier-1 Gogent
In coming months, I'll likely be putting the crazy disk RAID / SSD lust group to test with a real workload. Still perfecting the entire job, but I've rolled it out before and it's brutal and that's just merely me moving in. I'll even nice things and cap CPU. So we'll see how slow CPUs are + how laggy drives really are and how long something real takes, if they even let it complete.
I am pretty sure, there are going to be salty providers with unhappy drives when reality plays out in their containers.
Goal is to show numbers for what they are with generic benchmarks - useless shit indicator that is being gamed in mass.
HW cache is just one layer of aiding and abetting wonky / fraud numbers. There are others.
I am pretty sure, there are going to be salty providers with unhappy drives when reality plays out in their containers.
Goal is to show numbers for what they are with generic benchmarks - useless shit indicator that is being gamed in mass.
HW cache is just one layer of aiding and abetting wonky / fraud numbers. There are others.
WelltodoInformalCattle
New Member
You can't mention something like this without elaborating.It's also easy to win with nice DD's. The DD's are gamed, though, due to tunings in cards. I won't go into this much since that's for everyone else to debug, but I ran into a lot of it in recent weeks.
They took a few WEEKS last time didn't they?
DomainBop
Dormant VPSB Pathogen
Shilling? Didn't you wet your bed at one stage?
Software RAID 0 can also put up some nice numbers. I have a development server with 2 x Corsair Force GS in RAID0 that gets 763 MB/s dd test results under a full load...with my dd scores maybe I should buy that provider in a box kit and start my summer host RaidZeroVPS. Guaranteed top 10 showing in the Q2 poll.
