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
Vultr is prem.
Vultr.
it depends on a lot on the deal you get. Vultr $5 VM will beat a $5 regular price Virmach VPS... However, if you get a special Black Friday Virmach deal, you will crush Vultr easily... for $60/yr on Black Friday you could get 8+ cores, tons of storage, tons of bandwidth on 10gbit, extra ips...
horses for courses, as they say
both are reliable, I bake in some risk premium discount for VirMach (relying overmuch on ColoCrossing)
And I think Vultr (generally) has better network
But if I need mucho cheap ram, not so much single-core cpu performance - then a VirMach black friday deal is the bees knees:
$48.60/year 10 GB ram, 4 weak vcores, 30 GB SSD in Seattle
https://browser.geekbench.com/v4/cpu/14992890
compare with Vultr deal $42/year ($3.50 monthly) 512 MB ram, 1 okayish vcore, 20 GB SSD in San Jose
https://browser.geekbench.com/v4/cpu/14631814
neither of the above are standard pricing but I'm not usually one to pay retail
Benchmarks are one thing. Explain the use case.
indeed - as I said, "horses for courses"
for the specific instances I listed benches for:
teh VirMach for doing (occasional) full-text searches
and the little Vultr for a reliable jumphost / proxy
Sorry. Forgot to mention that I want to use LEMP to run my small blog using Wordpress.
Just get the $6 nvme from vultr.
Honestly, I don't think there is any difference in either case. If the price and specs are the same, Vultr probably has more resources and is probably going to provide better stability and network performance relative to Virmach (note that I am not saying Virmach is bad, just that comparing the two, Vultr probably has a slight edge in the grand scheme of things).
Thx guys!
Vultr for what you use in production. When your system down on weekends, it's likely you get support from Virmach. Look at this:
I am in the "Priority Queue" but I am still waiting for more than 30 hours for purchasing more bandwidth.
Vultr, in terms of support, panel and service. On top BGP is nice one you could use have for all their location for 0$ cost.
Don't put any prod on virmach, may be some dev or experimental or basic sites. Not that they are bad just that simply not meant for all users who expect something extra or support.
I've used both and they're great. I'd say vultr is premium and offers more features, while virmach offers some great specials, particularly on black friday.
But for wordpress / LEMP on that price range, go with the $6 nvme from vultr. Your site will run much faster with the nvme disk and that "high frequency" cpu + faster RAM.
100% Vultr
No question, Vultr.
VirMach suspends VPSes a lot. I use their standard plans and even some suspended aren’t being really used that intensively. In the end I canceled all services I had with them (had 10+ at one point) except one. Just had to spend the credit...
With Vultr I had a bug in my script that made one VPS of mine run at 60% or 100% (cannot recall which) for days until I fixed it (Well, mistakes happen!) although I had other VPSes that were relatively idling. But then they didn’t suspend my VPS, didn’t even send me an intimidating warning, and that is fortunate because I was running critical services on the VPS. It just reassures me that I can continue running the more critical services on Vultr. (I don’t plan to abuse Vultr’s resources! I use dedicated for that.)
People say Vultr is better than Virmacht, but give no stats. Support is not as important as the power and capabilities of VM, uptime and so on. Where are stats and benches? Btw, Vultr as far as I know doesn't accept debit/credit cards. Virmacht does?
Vultr accept CC, PP, BTC, Chinese
Support is important if you cant access your vps during weekends and you found out that the provider will be available by Monday.
For production my vote goes to Vultr. I dont recall submitting any support ticket for the past 6 years over 300 VMs. I have tickets but those are only for testing support. They do have weekends. And answers within hours not more than 24hrs. They do accept cc, prepaid, paypal.
VM performance is good. Network is good and better than most of providers. Depending on location. Hardware are really well maintained.
We also have ha infra in vultr, and i can say its much better there than others when doing private network connectivity.
I think vultr is not using ryzen yet or amd cpus. Only intel gold or other intel cpus.
@VirMach
We also use virmach, about more than 20 vms. Its just fine. Support is good also. We had never encountered bandwidth overage yet. So i dont know what will happen in case its weekend and we hit the bandwidth limit. Performance is just fine. But i think they still do not use nvme and new cpus.
I choose Virmach. @Virmach is awesome.
Could you please provide some benches? Monster bench and YABS will be good.
Please be more specific. You want benchmark of Virmach or Vultr? Which location/datacenter are you interested in?
You can get free credits on vultr when you signup and choose the dc you want to benchmark.
This ticket was actually potentially resolved in 30 minutes (I see a note on here, 22 minutes after creation) but the problem was that it was a dedicated server.
Then you replied to your ticket the next day, which bumped it down the queue and kicked it out of the "priority" flag that it was placed in. We've since fixed this to where priority flag gets auto re-flagged, and we actually also made it recently to where customers cannot reply to their own tickets because we had this issue. We use WHMCS, we sort by last reply. So customers who are in a rush reply thinking it somehow helps them when it's the opposite.
I do see after this, that it did take way too long to process your bandwidth request.
This was actually ColoCrossing being weird about the additional, with their staff saying they're unable to make the change. Usually it's added, and they update us that it's added, pretty much all our team members replied on the communication with them and I even called them out for it at the end. Definitely not acceptable.
We are definitely looking to make bandwidth additions more seamless in the future, but unfortunately I do not think this will be the case in terms of bandwidth upgrades for dedicated servers. I urge everyone to always do it in advance and otherwise assume the bandwidth you purchased is what you get with any upgrades to the server not guaranteed on an emergency basis (other than hardware replacement, but the team at Buffalo seems a little better with this one.) I'm still following up to see what happened here because I'm not sure why their team members said they do not have the ability to increase bandwidth and had to escalate it.
We implemented a priority queue for these since a while ago, as I mentioned above would it have been a virtual server instead of a dedicated server then it would have potentially been done in 30 minutes.
We are trying to make improvements here as well to automate it a bit.
There are definitely times where our support takes longer than I would like; we are still catching up to a backlog of tickets after a bad couple weeks. COVID, plus my internet keeps going out. Plus some other team member's internet has been out for several days. We've got rid of our outsourced support for the most part. It's difficult getting new hires in this climate, and we just had a bunch of migrations and conversions done too as well as some website software updates that ran into some problems. Times like this do happen with small teams.
You do get +500GB bandwidth and DDoS protection in Buffalo at no additional cost, plus Windows and +10GB for +$2/mo with us. And if you move onto the $10 pricepoint, we're one of the only providers to give 2 cores instead of 1 core.
We also have the Elite+ package still, and now that's KVM.
You could get this with 1GB RAM, 2 Cores, 30GB disk, and the same bandwidth for just $4 per month. So there's a higher CPU, but I suppose it could be more restrictive. More disk, but it is fair use technically (I mean if you resize your disk to 30GB and you're using it we don't care.)
Features, well, we're stuck with SolusVM right now so... it ends up going into this:
Then to address the abuse issue that pops up.
I am viewing anti-abuse logs for the previous day (24 hour period.) These are out of tens of thousands of VMS.
There are, for CPU:
For I/O:
I'll see if for I/O we can tweak it even further to shift the ones without warning into the ones that get warned first, but I think the issue is that for the immediate powerdown ones something actually went wrong very quickly and it had to be dealt with immediately.
I do see maybe one or two that should have been given more leniency.
These were both very low disk. One was clearly abusing. The other was borderline so it would have been better for it to receive more warnings. By borderline I mean it's still clear abuse, just that we could have afforded to give it more leniency anyway. I'll speak with the team see if we can reduce these immediate shutdowns even further. But to paint a picture, these were 5-10GB services doing 10,000-20,000 reads a second all day.
To find out what's better we need to see Virmach and Vultr both benchmarks. I'm interested in European location, but other locations would be good too.
Well +1 for @VirMach being alive here and for the constant improvements.
It's possible for it to go both ways depending on where you land. On a higher-end node with the CPU higher than we'd like it? 20% lower performance than Vultr's advertised rates. On a super old node we have with tiny cores? Also 20% lower. Our best compared to their best? About the same. Our best compared to their worst? Probably 30% better performance.
They are most likely expanding at a rapid rate and getting more new hardware than we are, and it's up to Skylake. They probably are big enough to where they can lease out the aging hardware to their partners or sub-brands. We keep it in the same brand.
I'm going to see if we can turn the super old hardware into its own cheaper sub-brand or just reserve it for specials.
Anyway, just for fun here's a benchmark of an empty dual E5-2687W v2 server.
https://browser.geekbench.com/v4/cpu/15505723
I got recently three shutdown for IO abuse (#518876, #215269, #442204).
It was a idling VM (Actually was serving a single file one times per months).
There where no evidence of compromised VM.
I monitored closely after that and never I saw any abnormal usage.
Support suggested a reinstall, I did it and shutdown are gone. Hetrixtool now report all usage constantly.
I'm a big Virmach fan, but I have a mixed feeling on this. All my VM work flawless and I feel bad for taking support time, specially on this really special deal. On the other side, to the best of my knowledge their where nothing wrong from what I can control and it was really hard to point to support that the issues may be outside of my control.
Vultr
IMO these two are not of one kind. If the project is serious, I’d choose Vultr; otherwise, Virmach to save some bucks.
And one major shortcoming for Virmach: no IPv6 support.
There are software bugs, too. If this happens again, find the app using the disk I/O so you're aware. This can lead to discovering typos or fat finger mistakes that made something run in a loop.