All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Digital Ocean Test ... Am I crazy?
Been looking around for some lots of bandwidth solutions on the Internets for hosting large files. I thought I would give DO a try, am I the only one that is seeing this crappy service?
wget freevps.us/downloads/bench.sh -O - -o /dev/null|bash
CPU model : QEMU Virtual CPU version 1.0
Number of cores : 1
CPU frequency : 2260.998 MHz
Total amount of ram : 496 MB
Total amount of swap : 0 MB
System uptime : 1 day, 3:30,
Download speed from CacheFly: 15.0MB/s
Download speed from Linode, Atlanta GA: 4.38MB/s
Download speed from Linode, Dallas, TX: 2.41MB/s
Download speed from Linode, Tokyo, JP: 1.51MB/s
Download speed from Linode, London, UK: 2.53MB/s
Download speed from Leaseweb, Haarlem, NL: 3.09MB/s
Download speed from Softlayer, Singapore: 1.76MB/s
Download speed from Softlayer, Seattle, WA: 2.97MB/s
Download speed from Softlayer, San Jose, CA: 3.17MB/s
Download speed from Softlayer, Washington, DC: 7.72MB/s
I/O speed : 80.7 MB/s
and here is my 128MB KVM at Hostignation
wget freevps.us/downloads/bench.sh -O - -o /dev/null|bash
CPU model : QEMU Virtual CPU version (cpu64-rhel6)
Number of cores : 1
CPU frequency : 3392.292 MHz
Total amount of ram : 121 MB
Total amount of swap : 236 MB
System uptime : 227 days, 15:15,
Download speed from CacheFly: 10.4MB/s
Download speed from Linode, Atlanta GA: 2.27MB/s
Download speed from Linode, Dallas, TX: 2.52MB/s
Download speed from Linode, Tokyo, JP: 1.27MB/s
Download speed from Linode, London, UK: 781KB/s
Download speed from Leaseweb, Haarlem, NL: 829KB/s
Download speed from Softlayer, Singapore: 604KB/s
Download speed from Softlayer, Seattle, WA: 2.45MB/s
Download speed from Softlayer, San Jose, CA: 7.59MB/s
Download speed from Softlayer, Washington, DC: 1.51MB/s
I/O speed : 108 MB/s
Comments
What makes you judge it as a "crappy" service, I presume the IO, which is a totally usable IO.
Are you on trial?
The IO is terrible and when sshed to it freezes every now and then.
@kamalnasser I was, I added 5$ from paypal so I assume not anymore.
Recreate the VM a few times to get a different node
No it's not. Are you continuously writing anything over 80mb/s? I presume not so it shouldn't matter. You can't base IO or even service of a simple dd.
Open up a ticket and ask them if you're on trial, existing trial droplets get buggy sometimes
@BradND are you serious? That would be awesome business answer please delete all your setup and create a new droplet on a new node LOL
Off topic : Interesting on CPU frequency : 2260.998 MHz, since it is up from their usual 1999.9Mhz
And Back to OP... so which is crap? you mean the IO?
On their Feedback site they state they just finish working on their DNS ( What is new? Do anyone know? ) and they will be working on storage next.
I cannot explain it, ssh seems to freeze for ~30 seconds every now and then. Sometimes the network just is slow as molasses and all, my old VM on Linode "seemed" faster and better then this node and its supposed to be SSD
80MB/sec I/O can't be that bad. Try an ioping to determine if it's disk issue, and check if there's packet loss.
Not sure how you can compare Linode to Digital Ocean, they both target different markets.
waaaaaaaaaahhh my IO is lower than someone else's; they must be shit!
You do not have to delete anything before recreating, you can make a snapshot and then create a new droplet from that snapshot (don't delete the old droplet yet, more chances that you end up on a different node that way).
@BoGs Then don't complain if you're not willing to listen to what people are saying. The I/O is fine. You can also use the snapshot feature.
How does this label it as "crappy"?
No, they target the same market. DO has many Linode crossovers using their services.
No, no. it's right opposite. I noticed that many hosts started to take advantage from stupidness of ordinary LEB readers who believe now that even 200 MB/s means bad performance.
@BoGs sequential I/O throughput speed result is not best measure of performance however it you judge barely on those numbers 80.7 MB/s is completely decent.
I think 30 is still usable for most things BUT, 80 MB on SSD raises some flags. How much is iowait ?
Yes you are crazy
Imho. big part of LET community lately became crazy regarding those numbers. We reached a point when people start to claim that even 200 MB/s isn't good enough and some hosts find a great niche to exploit this stupidness of average LEB customers. "You only need as much speed as you need" - as someone said in other DO thread.
Are you sure it's the IO and not the network? What location of theirs do you have? Have you tried to see if there is packet loss and latency issues?
We have massive amounts of bandwidth at vHostlayer each rack has a 10Gbit port, Accept our USA location.
If your complaint is IO, ok its not Ramnode, but seriously, were you really expecting Ramnode performance at that price? If your complaint is the bandwidth, realize that they are using very cheap bandwidth, all Tinet and Cogent, and most routes that use Tinet hand off at the closest Cogent peer. So, you are getting bulk rate Cogent bandwitdh, and your not gonna get 'the best speeds' from it all the time. This is the reason they are able to afford this 'Unlimited' or 'Unmetered' approach. Using better more premium routes would mean the cost of the service would have to go up quite a bit to be able to support higher quality bandwidth at the quantity they are using it. At least this is what I have gathered from my testing, feel free to correct me if I am wrong.
My 2 cents.
Cheers!
Works great for my simple needs of a small website for a couple clients I do some webdesign for. I honestly dont even use more than 5GB/mo. bandwidth and nothing I/O intensive at all.
So HDD I/O speeds are fine for my needs, the price is right for what you get.
Seriously?
I refuse to accept your USA location, if you try to force me, I will have you arrested. However if you meant 'except', this makes a lot more sense.
/Troll
I used to get 14-15 MB/s on one node, now I'm getting 220-290 on this new node I am on. So back up you droplet and restore from it, chances are you will land on a faster node.
Except our USA location..
Was just a suggestion as we host some big File hosts on our offshore locations.
I was stating the servers are located "offshore" Doesn't mean we ignore DMCAs etc but we have some big FileHosts that use our services.