I've been an OVH customer for a long time (years), and would never bet my company on such an offer.
OVH excels at offering lots of resources for a cheap price, but they are awful at support. Think of any horror story and I've either seen it happen firsthand or have heard about it at their forums. Examples:
- One-month back-and-forth with support about a clearly missbehaving hard drive (2MBps sequential transfer speed)
- A whole week of lost "virtual rack" (private lan) connectivity.
- Failover IPs (ip's that you can move from server to server, touted by OVH as a failover solution) stuck in netherland (routing to /dev/null because of some issue) for some days
- Extremely slow High-Availabily NAS speeds for a cuple of days. Afer those days I migrated off that service and never looked back.
At this point, I only hire dedicated servers from OVH. Furthermore, my strategy to deal with any issue is simply to rent a new server, test it thoroughly (important! OVH should do it, but they really don't), migrate everything, and don't renew the old one. In my own experience, this leads to far less downtime and headaches than trying to deal with OVH's support.
Many friends of mine have been or are OVH customers due to their low price and all of them say the support is just awful.
In my own experience, when I was consulting a company with a physical server on OVH and opened a ticket... Well, you know when you open a ticket, you don't know when they will be answering.
Their support is awful and their customer service is down right disgraceful.
I used to be a customer for almost 18 months. We got DDos'd twice. On both accounts, they didn't and said they couldnt do anything. The only response was to turn of the server.
Not only that, I tried to work with them on setting up a much bigger distributed environment. I gave them the exact specification of how to setup the network and clustering of the servers. They got it wrong, on multiple times. The Sales Manager who I was working with said on many occasions that I would NOT be billed until the specification was setup correctly and running.
Guess what, we were billed. The setup was never fully completed. I wasn't allowed to talk to the 4th line (network) engineers who were setting up the infrastructure.
In the end, I complained to the top level management with the timeline of events. I saved every email and when read through a chronological timeline. It was clear the sales person was inept and technical didn't do what was required. I never heard anything back.
What was worse, is that we got a partial refund. We lost $2k in total. I told my partner to go back to the bank and demand a charge back. But he had no spine and said we'd eat it.
Never again will I go with OVH and I tell everyone I come across not to bother. Sure they prices are cheap in the beginning, but their support is non-existent and their engineers awful.
If you call their phone support, they’re far better. They actually reply instantly, help you step by step through things, are even able to provide you with information right away, and fix things.
I’ve only ever used their phone support, so I can’t say anything about their online support, though.
Hm, then maybe it was just a coincidence, or it depends on language (I used the support in German).
Tbh, I don’t expect support from them anyway – their dedicated systems are so cheap, for the cost of hosters which provide support I can rent a handful of systems at OVH and have the same reliability (and higher peak performance).
Please note that all RunAbove stuff is Beta and will be discontinued at some point in the future and either integrated in OVH.com or completely dropped.
Had a service running on RunAbove Cloud, and suddendly they pulled the plug with 1 month notice - and nowhere on their page was something that hinted on that.
It either wasn't in the past or was significantly less clear in the past. I recall using runabove and within runabove there were beta trials but it seemed like a generally available service for the other parts. I was quite surprised coming back to it that I couldn't spin up a box there like I had done very shortly beforehand. I received no email about the change.
Here's a quote from their about page previously:
> RunAbove aims to provide a cloud infrastructure powerful and transparent, for developers all over the world so they can focus on their work by relying on unshakable resources.
Unshakable resources indeed.
It switched over about September/October last year.
While I love seeing more PostgreSQL offerings in the cloud, I don't understand why they call it PaaS. It's a DbaaS, but a PaaS... a PaaS is really an application API that is managed by the cloud provider, and takes care of stuff like scaling it. Like Google App Engine. But calling a database a PaaS is a little bit bending the term, IMHO.
> [an application API that is] managed by the cloud provider, and takes care of stuff like scaling it
A system like DynamoDB, that exposes no concept of database cluster size, would be a PaaS, then? Still sounds off, to me. "PaaS" just isn't a word that applies to a service that doesn't execute your code.
Something that's almost entirely a database on the backend could still be a PaaS if it runs your code, though—like Parse.
Friends dont let friends use OVH, I cant reiterate this enough. They've long and sometimes dramatic history of broken services, undelivered paid services (like months long queues for paid upfront servers not being delivered), lack of communication with customers, truly horrible support all of which I had the back luck to experience myself, so do yourself a favor and just avoid them at any cost.
Interesting .... and the prices seems to be low,
but they run postgresql 9.4 which is a little outdated.
and there is no details about what you'll be able to do / not do (like installing extensions, or other similar things)
The VPS isn't fault tolerant or managed. I am assuming this is? I actually asked them on IRC if they could launch MySQL, PSQL aaS less than a month ago. Turn around from the RunAbove team has been super fast.
Cheaper, but does it also provide value? For non-critical things, OVH provide relatively good value for money. For critical things, I wouldn't touch them with a ten foot pole.
OVH is a dysfunctional company that somehow turns a profit.
We almost lost our company's domain name twice due to their incompetence.
We paid OVH for the renewal of our company's domain name several weeks before its expiration. This domain hosted our google apps for business and more. We received an invoice and bank statements for that. OVH never renewed with the registry.
We only got to know about this because of a standard email warning from the country's registry (not OVH as a registrar) itself about quarantaine and expiry!
We spent over eight hours over several days talking to OVH. I have been able to verify that the problem was on the side of OVH only, and not on the registry's side at all.
In the end, we made a direct payment to the registry to make sure the domain was not deleted.
I've been using ovh (proper OVH brand) for years and the only issues I've had recently was a service without an SLA which I foolishly used without a failover in a different dc (load balancing ips had an outage for ~5h recently)
Other than that they've been top notch, support is slow, but the non-human things are working as advertised
To offer an anecdote to counter an anecdote, I'm using OVH proper as well as SoYouStart, one of their "developer" brands and I've had zero issues with either.
Actually compose.io is HA (Highly Available) while the runabove isn't, so thats just not a good comparsion.
It would roughly compare to AWS RDS (just that you can't) kill the instance after a half month and only pay half the month. Then AWS would've been on 15 USD ~ 13 Euro with 1 GB Memory and 16 GB Disk, compared to: paasdb-1g that's a lot more, but RDS is really solid even on a Single Node, so it's hard to say if you get the same quality for just 6 Euro. It could be true, but it doesn't need to be true.
Actually AWS RDS offers High Availability[1] at a higher price from their single AZ deployments. I've been using it for quite awhile with no issues so far. Compose looks much more expensive than RDS with multi-AZ.
we are using that as well however this post was about single nodes so compose doesn't offer a single node and I wanted to point it out that there is AWS RDS which offers a single node.
I believe OVH offering will be cheaper, and hourly billing but only available in Europe. While compose.io's offering have more locations (US, Europe, Asia, etc).
It is good to see more and more PostgreSQL as Service offering.
As some said, OVH has datacenters in Quebec, Canada.
They plan to open several new ones in the coming months in the US (I think California has been officially annouced) and others countries are planned.
They raised several hundreds millons of syndicated loans in 2014/2015, to build additional capacity in North America and Europe.
Rumors[1] of potential asian & australian locations also abound, but no clear sign of it yet.
I'm mainly interested in EU SaaS (because being in the EU). For EU companies - the verdict is still out - EU locations of US companies might not be enough. One reason why Microsoft lets Azure in Germany be run by Dt. Telekom.
Many of our customers start including request to the effect of "hosting in the EU, preferably operated by a EU based company".
AFAIK, one of the reasons OVH had for not setting shop in the US a few years back, but choosing Canada instead was to not expose its flank to possible actions by US agencies. They even feared "contamination" to their EU based datacenters.
I guess either the law has changed or their lawyers found their way through the arcanes of international law, since they are now officialy planning to open a datacenter in the US.
As others have noted elsewhere in this thread, all RunAbove products are 'in beta', and they have a tendency to terminate these services with very little notice given.
I wouldn't rely on them for anything that needs to stay up.
OVH excels at offering lots of resources for a cheap price, but they are awful at support. Think of any horror story and I've either seen it happen firsthand or have heard about it at their forums. Examples:
- One-month back-and-forth with support about a clearly missbehaving hard drive (2MBps sequential transfer speed)
- A whole week of lost "virtual rack" (private lan) connectivity.
- Failover IPs (ip's that you can move from server to server, touted by OVH as a failover solution) stuck in netherland (routing to /dev/null because of some issue) for some days
- Extremely slow High-Availabily NAS speeds for a cuple of days. Afer those days I migrated off that service and never looked back.
At this point, I only hire dedicated servers from OVH. Furthermore, my strategy to deal with any issue is simply to rent a new server, test it thoroughly (important! OVH should do it, but they really don't), migrate everything, and don't renew the old one. In my own experience, this leads to far less downtime and headaches than trying to deal with OVH's support.