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
here my DD
dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync dd: writingtest': Disk quota exceeded10781+0 records in
10780+0 records out
706490368 bytes (706 MB) copied, 1.4512 s, 487 MB/s`
The price is irrelevant when talking about performance. If there were no performance limitations listed in the promo, there should be no performance limitations in the service.
This is a common error made when talking about hosts/services. If I sign up with a host, for example, that guarantees 99.99% uptime, then it doesn't matter if I'm paying $1/month or $50/month, I'm paying for 99.99% uptime. The same goes for 24/7 support and 15 minute tickets. If that's what's advertised, then that's what should be expected.
If it's not reasonable for a host to offer high quality plans at low prices, then that's fine. Either don't offer the promos, or set the terms of the promo appropriately. That is the only way to do honest business.
Oh come on. $7 = $6.50 after PayPal fees = $0.54 / month!
@jemaltz
I think I could agree 100% on uptime and support. However I/O performance? Really?? As long as the box is working for an entire year at $6.50 (after PP) is amazing on it own with the specs provided.
...and?
I didn't tell them to have the offer, and I certainly didn't tell them to make the offer under the umbrella terms that they advertise!
Honestly, I'm really fine with this attitude, as long as it is stated in the terms of the offer. If it is not, then as a customer, I will go by the rest of your marketing literature. It's not very nice for a host to use uptime/support/performance guarantees to win my business, and then complain that I'm only paying $x (a price that they set!) when I hold them to it.
Again, I really have no problem with lower support/performance/uptime levels on cheap deals as long as that's in line with the terms of those deals/services.
@ jemaltz > jemaltz said: uptime/support/
Maybe I'm missing something. uptime/support = Spot on 100%. Where did he make performance promises?
When some people have 250-400 MB/sec I/O and others only 12 MB/sec, and they're all on the same plan, and paying the same, and the latter group of people is just told to suck it up and don't complain because it's "so cheap"? You really have to be a bit damaged in the head (maybe that comes with becoming a host), if you don't see anything wrong with this.
He didn't state that these plans would have lower performance than any other plans, so performance should be equivalent; or at least meet the advertised specs on the website.
While vague, it does say (as of today, at least):
If I didn't order a VPS advertised as having a 90+% performance reduction, then I'll be a bit unhappy if that's what I receive!
Unless your use case is wget-ing linux isos at gigabit speeds, I'd also check out your fio/ioping scores.
My two birthday-specials are still benching around 7000 IOPS minimum. (used to be 22k-28k on fresh host)
7k is insane value for $7/year.
This is one snappy container with no iowait.
If you're happy to demand super duper I/O for $0.54 / month, go for it. Just don't start Sweat Shop Node or something.
Hold them to what? Do they have I/O performance guarantees?
Oh, the contrast. I wasn't looking at it that way. You have a point there, but I wasn't thinking about jealousy. I meant noticeability as in if this thread hadn't appeared, would many in the lower I/O set have even noticed the 'lower performance'?
It depends a lot on the nature of this slowness, i.e. is it hard-capped at least with the benefit of a low and stable ioping at 0.1ms, or is this just because it's oversold to the brim, with ioping all over the place up to 300-500ms or more.
In any case, they absolutely have the right to complain, and "but it's cheap" is not an acceptable response. Not only because another arbitrary group of customers at the same host randomly got a 30x better performing package, but also because they probably wouldn't want to buy this deal if they knew of the performance they get in the first place.
There's a lot of $12-$15/year deals which will shred not only the "lower" but also the "higher" performing groups in terms of actual performance, and compared to $7/year that's not a lot of money to add for the improvement you can get.
Mine:
Is the throughput consistently "slow"? One (or a couple of) runs of dd is hardly telling a useful story.
nevermind
My I/O is a lot better, see Screenshot. I'm on node 5 btw.
Too bad the top result doesn't work on windows server editions... see why I asked what people USE?
So, I've had some weird experiences with a Windows VPS and after reading the thread decided to test its I/O. How does this look, I'm not sure what to make of it.
Same box different time of day, bit better, like I said perfect for what I'm doing.
If I were to write to the disk a lot I doubt I'd notice too much smaller files run great..
Honestly how often are you going to write a 1GB file ?
Looks exactly like a counterpart of "12 MB/sec dd", or what do you need an explanation of, what does "Sequential write: 11.57 MB/sec" mean on your screenshot?
Every day m8 and twice on Sundays, also you seem to think
conv=fdatasyncwas just some irrelevant detail on the command lines everybody else is using.Lol! You're so quick to judge "m8" (that's actually my phone's model..) I seem to of missed that part of the command when I copied it.. wow..
You need to chill out a bit it was just a general question you didn't have to answer like a total twat...
dd if=/dev/zero of=test bs=8k count=16k conv=fdatasync 16384+0 records in 16384+0 records out 134217728 bytes (134 MB) copied, 0.477766 s, 281 MB/sNode8.
From a $7/year VPS on dalyearly1, the node most of you are bitching about:
It's also nice and responsive, and not showing a lot of iowait.
Who gives a crap about sequential write speed on a VPS? Oh noes, it'll take 9 minutes and 57 seconds to fill your whole 7 GB disk allocation, time to open a thread.
A VPS node's workload is almost entirely random read/write. The "dd test" can be indicative of an issue but it takes 10 seconds to check it out with some other measure, either another synthetic benchmark like ioping, or seeing how it's performing with its real actual workload (e.g. metrics like page load times).
But everyone would rather wave their dicks around complaining about their $0.58/month idle VPS's performance. And then you wonder why certain providers don't post here any more.
Half of you guys who got 200MB/s+ are probably on Dime nodes for their "regular" plans with RAID10, Oktay did a special on raid 1 e3s for $7/yr.
How about actual clients from that promo on nodes dalyearlyXXX post instead.
Or... maybe... I'm asking about the difference in results between 512k and 4k, if it is normal because THAT's something that is not obvious to me as I am not familiar with the fine art of judging performance metrics.
Something that a well meaning reader would notice. You on the other hand, fail to qualify as such and so we end up with your answer.
This is from my $7 box:
> dd if=/dev/zero of=test bs=8k count=16k conv=fdatasync 16384+0 records in 16384+0 records out 134217728 bytes (134 MB) copied, 0.44493 s, 302 MB/si too have one.. really happy with it..
What are you guys doing with your yearly boxes?
I installed XFCE and xrdp on it, and use it as a remote desktop
@Gunter I am running a slave for my Powerdns / Mysql setup on it
Mostly idling, what else??!