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.
Can Crunchbits be used for Production?
They have been around for a while and have been voted one of the top hosts here, but my VPS has been very unstable lately. I received an email about their maintenance
Maintenance Window:
Start Time: 2024-09-02 0:30 UTC
Expected End Time: 2024-09-02 4:30 UTC
They expected around 4 hours but here is what my VPS server has been like for the past few days.
Downtime: 16 hr 4 min
Noticed at: 2024-10-02 22:25:27 (UTC+00:00)
and
Downtime: 12 hr 5 min
Noticed at: 2024-10-01 06:55:48 (UTC+00:00)
So my question is, do you use them for production?
Comments
The fact you have to ask this answers the question.
Generally, you're looking for a SLA.
For what it's worth I use them for production
If it's affecting critical infrastructure, make sure to have a backup you can fail over to
Their network from NA to APAC can be slightly iffy sometimes.
I'm at 99.95% uptime with my dedi for the year (I didn't bother with putting the monitoring agent in maintenance mode, so that's a tad skewed) but I use it all day, every day and it's been extremely reliable. One of the handful of hosts here I would actually trust for something in production.
If anything does go wrong, Eric's got a crew of super talented guys that keep the train-a-chuggin'. No lack of experience there.
As @MrRadic said though, always good to have a backup for anything in production. Usually want to get those at a geographically distant location from your main server so if something like a fiber gets cut (happens all the time) your backup server won't be affected.
The past couple of days is not a normal indicator. The "maintenance" was physically packing everything up and moving from Allentown to Valley Forge, PA. Relocating from one datacenter to another is bound to have unforeseen issues.
I really do apologize, we definitely had some issues with notifications (mainly around the converted WA-->PA Yearly and existing Yearly product stack in PA which wasn't getting tagged as a "PA" product for notifications) and I definitely could have handled it better. To be fair to us as well there were 2 separate maintenance windows in PA that you're conflating into one. The initial physical move, and then a small subset of hypervisors I elected to get live with reduced network capacity so my team (and myself) could get some sleep after 24 hours non-stop. We then required a second maintenance window on those 7 hypervisors to upgrade their network and bring them back up with full capacity. There were a few hiccups as well, such as some VM's needing to be restarted via the control panel (and not CLI) to get back online and a few machines having physical issues crop up that needed to be troubleshot and repaired.
We rebuilt the entire PA network from scratch (hardware, upstreams, and more upgrades incoming) and also moved approximately 1 hour down the road to a different facility with the same company due to needing the cage space for growth. It's been "in the works" for approximately 5-6 months and it was a massive PITA trying to juggle everything. Getting so many people on the same page and then dealing with the inevitable "who didn't get the memo?" and "what broke from the 1 hour ride?" issues. I believe we're good, though I probably won't officially do a long form follow-up until the weekend.
/pity party off
As far as using us for production: My goal is to make that the case. That was a big reason for this move. We'll have significantly increased capacity and redundancy on our PA network. Will it be perfect? No--hardware can still break. Hypervisor hardware may need firmware updates/etc that require short downtimes. My goal is to do the absolute best we can for the lowest price we can offer while staying solvent and sane.
This is very subjective. Anything can be used in production. But, for major uptime critical things I use Hetzner cloud or BuyVM at the moment.
But -- I have had 2 hosts at Crunchbits for a while now - one for Proxmox backup server replica, and one for an inbound/outbound proxy which is redundant with another provider.
I intend to and have thought greatly into moving my primary stuff I have in various cloud services to Crunchbits and using someone else as secondary.
It would be awesome to be able to provision quickly, maybe be able to add vcpu/memory/disk space as needed, hourly billing to build and destroy at will for testing deployments and such. I do like what they have going on and their ownership of their hardware, tending their own network/AS and not just getting some blend drops from the DC they colo in. The IP ranges are actually not obliterated by fraud and bad behavior.
If they offered a service to colo single servers (1-2U, I'd probably colo an extra box I have with them since I live in the same general area in PA that they colo in.
Generally it feels like they are trying hard and improving as they go. One of the few not scammy providers on here.
I'm no expert, but for production wouldn't it be best to work with Docker Swarm/Kubernetes and implementing load balancers (something like nginx/traefik/caddy and cloudflare load balancer), then deploy on multiple providers globally, this should mitigate any issues like this from happening for most use cases.
and obviously backing everything up in multiple places.
No, only for idling.
/s
I have servers with them in both locations for my production setup.
Redundant though.
Yes.
Yes. with backup of course.
I'd say for many small businesses, scouting for a good reliable host for their production stuff makes sense, but clustering node apps, databases, caching etc might not make sense to avoid maybe an hour of downtime when redeploying every ~3-5 years when something goes tits up.
You are one of my fav providers here; but just a small friendly suggestion: a little more communication should be done with your clients, as I have not received any notification about this maintenance. (nothing in my junk folder as well)
You said more maintenance is due for PA. Do you think you would need more time to get everything back to normal?
Of course, I have backups in place and have moved everything away for now and waiting for the maintenance to finish (fully) to move my sites back.
For production you go with big 3. If shit goes sideways, at least you won’t be alone as 50% of internet will go puff.
Everything else - hobby, side projects. Strictly.
I use them for production. Only reason I'm not buying from them anymore is that I've upgraded to running stuff via colocation.
@Levi who is the big three you referr to?
aws, azure, gcp
AWS, Azure & GCP is not needed for production. They're needed when you're using their managed services. There's plenty of stable providers that are not one of those.
I've got multiple Crunchbits VPSes, and I'd say they're production ready. (I've been a customer for about a year, and I've only used the Washington location, fwiw.)
Downtime has been practically non-existent and for the prices I've gotten, I think it's a good/fair deal. — I have cheaper VPSes elsewhere, and they'll just randomly go down for hours/days.
m y i m love crunchy
I have 2 VMs, in PA and WA. Barely any issues and in prod. That doesn't mean I don't have fallback VMs with the exact same setup.
In WA, not a single host outage, here's my uptime. Only network maintenances and such.
PA I lost my uptime because they recently had some network maintenace/outage (can't really remember) and I had to reboot the VM to get it back up.
the server itself? prem, super prem
network from their server to APAC? meh
I'm not a customer but probably will be due to how you responded here, well done sir.
I have a good experience with Crunchbits, with two VPS from their Spokane location. I use one of them for "production" without major issues. Both VM's uptime is great. Their Support (via ticket) fast and friendly. Previously I detected some short network outages, anomaly only (in the last 3-4 months: about 2-50 mins). Overall, I can recommend Them.
What's your "Production" used for?
It doesn’t matter where you go for production. You always back up your data and have a DR strategy if it’s truly critical.
Otherwise, I feel safer with Crunchy than a big cloud provider secretly training its AI on my Services.
Have been using Crunchy for more than an year... Renewed all my servers for another year without any concern. In fact they had 2-3 Fibercut issues from their upstream but Crunchbits immediately switched all their servers to their backup network with pretty much 0 downtime. Support is da Best! That's what I call professionalism.
Rest for production purpose, it just depends on your requirement. Some people feel extra safe by paying more premium & going with Azure, AWS etc. For me Crunchbits was Perfect!
I do. But just as a node in k8s which hasn't been down yet
I actually have VPS in PA location, according to
https://uptime.crunchbits.net/status/public
The maintaince has not finished yet, so I'm waiting for an email from @crunchbits to move everything back there.