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
I don't know but in my mind @dataforest is ready for production.
I highly recommend @Calin if you just open a support ticket at ihostart you can connect with him and tell him your requirements
And if I may state the obvious, be advised that crypto is volatile is your money is not always safe……
Or Netcup RS?
I have good Experience on Using @ShockHosting & @dustinc @LiteServer @at Production.
I use a netcup VDS for production/business since years. I'm transfering my storage from LiteServer (which are decent) to @host_c (faster and a lot cheaper) and will use @Hybula for production/business as well.
Two important remarks:
That gives you way better safety/availability than this provider vs. that provider. And anyway be sure that the providers you use actually and really know the technical side of their operations and do not just run a sales operation with a glitzy shop front.
And yes, that also means that I strictly avoid resellers of any kind but always deal with the actual provider running the operations.
I know for a fact that Hybula are serious professionals, that host_c is a serious professional and that my netcup VDS has been very reliable since years. I want to talk to engineers and not to marketing or sales people and I want experienced engineers to run the operation.
I'm planning to use @ProHosting24 for my latest prod. I hope that they could always be this good or even better in the next.
A managed hosting provider might guarantee you a better service regarding support and fix your issues quicker, but that doesn't mean its automatically high availability and its not, if you need "100%" uptime, its not with managed hosting that you get there, but with server redundancy.
Now of course this is a matter of cost vs reward, does your business loose enough money if its down 0.5% of the time in a year, that it makes sense to mitigate the issue?
I'm currently using @xTom @NDTN @labze and @VPSSLIM for production.
The quality has been increasing for years, and there are more years to come
I wouldn't exactly say my stuff is "production", because while things are "available", I still don't have a ton of traffic because my product (still) hasn't launched beyond a few beta testers.
@fiberstate's comment about redundancy is key. I have a bunch of low-end providers that haproxy through to a bunch of API workers, some are VMs on my dedi, some are on other low-end hosts. They all have local sqlite databases and I use litefs to replicate those back to a central server (again another VM on my dedi), which merges those changes into a central DB which is then replicated back to all the workers.
At the moment, the haproxy instances are just listed as multiple A addresses for the DNS records, although at some point I plan to migrate that to geo-based DNS. But any of the haproxy instances or workers can go down, and it al just works. If one of the haproxy instances go down, a new visitor might have a couple of seconds delay until it tries one that works.
But basically, anything can fail and as long as I have at least one of each type of server still running, my site will still be accessible, albeit possibly at degraded performance (if for some reason a haproxy can't talk to a nearby worker).
The actual dedi where important stuff happens isn't really involved at all. Actually, I do have VMs running haproxy and the API workers on that too, but I've never bothered listing that IP in the DNS, but I could do should I need it urgently. But I'd rather not have any publicly accessible service on it really, because that might invite direct hacking attempts - even SSH access is filtered to one other IP over public internet or over wireguard.
I've experienced very little downtime from individual providers in general, although there have been a couple of times when a particular one has had an unusual long downtime. It was actually quite nice seeing that the failover worked as intended. For this, it's best to have several different providers in different datacentres or even countries, so multiple cheap low-end providers works really well for this.
All my stuff hosted on VPS is personal hobbies, services, (encrypted) backups etc. So yeah, I’m using them for production in the sense that these are my live instances, but there’s zero commercial impact for any of these and I derive precisely zero revenue from what I host. If they go down I lose my personal time but nothing else.
If a provider I’m with dead pools I probably end up financially better off actually, as it’s one less bill I’ll be paying and I’d probably just drop the impacted service on an idling VPS.
My comment was not to bash your setup. For a fairly large income generating site/blog/app, you have a good setup idea. I am keeping a simple formula nowadays, when my blog generates more than 20 dollars a day, then I have to move it from any LET host to any managed host with HA setup .
EDIT : Wordpress.com business plan has HA setup with server/geolocation redundancy ,for example .
I didn't think you were bashing on my setup
, don't worry.
As a sidejob i keep a similar redundancy setup for a company that generates 6 million a year.
And of course, the company i work for as a day job, which has a 1 billion a year revenue, keeps everything on AWS/Azure/GCP
Decide:
Priority status hosting with a let host
Being one of the many clients of large concerns.
The company I work for has been colocating with Hetzner for over a decade now, Hetzner has been growing and the company I work for did as well, just last year we upgraded our servers with a quarter million euros worth of hardware. So, yeah, it really depends on your definition of "production" and what you're looking for in a provider.
Using 3 Hetzner servers hosting 10+ sites in production for 4, 2, 1 years respectively, zero downtime till now.
Running 1 RackNerd server hosting 1 (not so important) site in production for about a year. My uptime tracker sometimes reports that the site went down for a minute or two, about once a month, but other than that it's been decent for $1.5/month.
Bought a couple of NetCup servers (one VPS, one RS) this Black Friday and planning to host some sites there soon.
I use Dataforest/Avoro (VPS, for one year now) and Shock Hosting (Shared Web Hosting, for 8 months now) and I am very happy with the services. Downtime is hardly seen on both services, if any... I use them for a public-accessible, legal thing but I don't generate revenue from that - I would still call it production use.
Not using anything for production, but if I did use anything for production it would probably be @alwyzon
I use Fiberstate (@fiberstate), Servarica (@servarica_hani), and OVH for production.
Currently in use :
Clouvider – incredible network & uptime 🔥
AlphaVPS – very stable; I store a lot of data here.
BuyVM/🏗️ – excellent performance 📈
Hetzner – perfect mix of performance/$ & reliability.
OVH – DDoS protection. * chef's kiss *
MXroute – works better than Mailgun, etc., for transactional.
Previously Used in Production (Still Recommend):
Worldstream
IncogNET
PulsedMedia
EthernetServers
Yes ... Contabo here!
We were mentioned a few times so if you have any questions, feel free to ask. We have some servers on the more affordable end that might suit you and have been around for about 18 years!
I must say, among the providers talked about here, I only really feel confident using Hetzner for actual work projects. For anything else, I’d probably stick to personal or hobby use, nothing too serious.
All the best from BuyVM @Francisco and @NameCrane
Awsome.
And OVH & Netcup?
Hetzner, UltraVPS, MXRoute.
I asked this question too a little while ago, some good answers there
https://lowendtalk.com/discussion/199925/which-let-providers-do-you-trust-for-production/p1
Hetzner, but I wouldn't trust most of the providers on here with important workloads.
I don't have much experience with them myself
Thanks everyone. For production I mean any product you wan the world to access to. Not personal hobbies or backups.
Seems like Hetzner was the most mentioned. I wish they had a smaller plan cause I wanna start very little as grow as needed.
I mean I don't need 4GB of RAM for now.
The problem isn't "is this or that provider advertising on LET stable enough?", but it is a safe enough place for "production" data. If the provider disappear and its disks are destroyed, that's fine as long as you have a good backup strategy or some kind of redundancy. Having your service running on several cheap hosts will likely be more stable and possibly cheaper than running on one expensive server BUT can you trust these cheap hosts with your (customer's) data?
Staying with reputables providers (Hetzner/OVH/Netcup/Scaleway for example), you limit the risk. Their nodes are likely kept up to date. Their panels and internal systems are likely secure. You can't be 100% sure than shit won't hit the fan, but it's less likely than say if your data is hosted in the basement of a broke "sister"-addicted guy who installed a panel he didn't write and can't understand, probably doesn't have the funds to pay for a good livepatching solution (and doesn't care), might leave its servers IMPI wide open, etc....
IMO while you could reliably run your main workloads on cheap hosts, it's not necessarily wise to do so. If you want to use them, it's probably better if it's for encrypted backups (if your data is somewhat private). These hosts might be stable enough for production stuff, and even more so if you use two or more of them, but that's not the only thing to take into account...