181–190 of 289 posts
Earlier quoted context omitted.
Quality decreased? How? Please elaborate.
I would love to hear why as well! Our quality and stability have significantly improved over the last year and as a co-founder of DO, I would like to understand what specific issues you faced. Feel free to shoot me an email directly (mitch@digitalocean.com) to take this conversation offline.
Benchmarking using wrk the smallest linode (1024 now 2048) serving a page from an untuned Rails application using nginx/passenger getting almost no other traffic. Hard to compare of course given the various other factors, but produced slightly lower performance after the upgrade. Serving a page from nginx directly (no Rails) had no appreciable difference in performance, I guess the Rails web serving is more vCPU boun…
Earlier quoted context omitted.
Containers, e.g. through Docker, are good for that kind of thing.
Not with the tiny amounts of memory offered at that price. 2GB is very easy to burn through in short order. Anything less than 8 feels really claustrophobic. I've always needed memory more than disk. Really wish I could pay some of these providers just for a one off increase of a few extra gigs of memory without tiering up which includes an entirely higher level of storage that I will not ever use.
I wish Linode (or anyone else other than Amazon) provides a reasonable Plan[1] with GPUs on them. [1]: Amazon charges $2 an hour thats about $1500 a month.
Earlier quoted context omitted.
Looks like library.linode.com has the same format, and is still the same fantastic resource. I'm embarrassed to say that once upon a time I used library.linode.com to help walk me through my first Linux server setup... running on Digital Ocean. Yes, I felt like a complete douchebag. I wanted to use Linode but I was one of those $5/month people. However it left me with a sense of obligation to give Linode business som…
Yeah, the library is really excellent - easy to follow and very extensive. It's smart of them to do paid crowdsourcing to create the articles. It's cool of Linode to make the library available to everyone, not just customers.
Earlier quoted context omitted.
They introduced hourly billing like a week ago :)
As long as you have a Linode provisioned, you'll be charged for it even if you're not using it. The hourly billing is just a simplification of their old system, where they would charge a prorated amount when a server is created based on how far into the month you are and issue a credit when a server is deleted based on how far from the end of the month you are. The net financial impact for a given server setup is abo…
Earlier quoted context omitted.
>I just have one server that I use to put stuff up on the internet and host a few services. And Linode treats that one solitary server better than any other VPS host in the world. I use DO for this and have used many others in the past and I have never had any problems with DO...
Anecdotally, I get notifications of maintenance outages every month with DO in NYC1 and NYC2. These are exceedingly, exceedingly rare with Linode's Atlanta DC. Also, stolen CPU on my DO boxes is super high, whereas this is very rare on Linode. You have less neighbors.
I have put this down to the DO SSDs. It will be interesting to see if this is true. I've been considering migrating from Lindoe to DO because they were faster and cheaper.
I've had 2 DO VPS for about 6 months, and never had a problem with them.
Why do I pay Linode $20/month instead of paying DO $5/month(1)? Because Linode treats their servers like kittens (upgrades, addons/options, support), and DO treats their servers like cattle. There's nothing wrong with the cattle model of managing servers. But I'm not using Chef or Puppet, I just have one server that I use to put stuff up on the internet and host a few services. And Linode treats that one solitary ser…
DO?
Earlier quoted context omitted.
Makes a lot of sense. If your hosts have, say, dual 4-way CPUs, and you're giving your VMs 8 vCPUs, then a single VM can execute per clock cycle since a VM needs all of it's vCPUs made available to the guest OS. With 8 VMs, that means one VM is executing every 8th clock cycle. If you "downgrade" those VMs to 2 vCPUs, then 4 VMs can execute per clock cycle, and that VM can now execute every other clock cycle instead o…
If your hosts have, say, dual 4-way CPUs, and you're giving your VMs 8 vCPUs, then a single VM can execute per clock cycle since a VM needs all of it's vCPUs made available to the guest OS. This seems wrong to me. We're talking about virtualization; technologically, it is absolutely feasibly to have only one physical CPU running a VM even if the VM sees multiple virtual CPUs. And it seems like having this capability…
Gabe spells it out pretty well here: http://www.gabesvirtualworld.com/how-too-many-vcpus-can-nega...
In a previous environment of mine, we ran dual socket, 4-way CPUs. VMs were configured with a mix of 1, 2, and 4 vCPU VMs. Our VMs appeared slow, especially on our 4-way systems. However, CPU utilization was low. More digging revealed that our co-stop values were high, meaning that the system couldn't schedule execution time effectively, meaning the VM had to sit in a READY state, which kept CPU utilization low.
Our first fix was to rebalance our cluster, so that 1 and 2 vCPU VMs were relegated to their own set of hosts, and our 4 vCPU VMs executed on their own set. Instantly, out co-stop values dropped, and CPU utilization rates went up...they were now doing work!
VMware briefly talks about the issue here: http://kb.vmware.com/selfservice/search.do?cmd=displayKC&doc...
If you'd really like the nuts and bolts of it, then those can be found here: https://www.vmware.com/files/pdf/techpaper/VMware-vSphere-CP...
The VMware co-scheduler has improved over the years, but I still read (The "Mastering vSphere 5.5" book by Scott Lowe carries a warning on this as well) that carefully balancing vCPUs is a must in a VMware environment. (Again, I don't believe Linode uses VMware, so I can't say with any certainty that KVM or Xen exhibit this behavior.)
So why can't we run 8 vCPUs on one physical one? Because while they're virtual to some extent, they're not completely abstracted. Anytime the hypervisor has to perform a translation between the guest OS and the host, a performance penalty is incured. So while the hypervisor may abstract scheduling, it reveals as much of the physical CPU to the guest VM as possible. Here's a little blurb from an older VMware manual explaining a bit of the difference: http://pubs.vmware.com/vsphere-4-esx-vcenter/index.jsp?topic...
CPU virtualization =/= emulation
For this reason, (again, at least in a VMware environment) we can't give a VM more vCPUs than exist pCPUs to align them to.
Hope this helps!
Earlier quoted context omitted.
Upgrades/downgrades requires transferring all your data from one server to another - so that means downtime. The bigger the server, the more downtime in transferring. As for backups, Linode backups seem to get a bad reputation. They do take a snapshot of a running instance - but reports from the forums indicate that they regularly fail, leaving you with a missing backup at some points. The only way of restoring a bac…
upgrades for ram and storage does not have to have any downtime. same for snapshots, through LVM/ZFS, they can be nearly instant in most cases. it just seems like DO have not been able to figure this out for whatever reasons.