Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!


Should I use Scaleway for production websites ?
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.

Should I use Scaleway for production websites ?

The VM packages at Scaleway are very good and its price is cheap. Should I use them for production websites ?

Comments

  • Have you considered https://www.online.net/en/dedicated-server/dedibox-power8 ?

    PS: Depends on how important are those websites, required availability, etc. Generally speaking, it should be OK.

  • NekkiNekki Veteran

    chrisfe said: The VM packages at Scaleway are very good and its price is cheap. Should I use them for production websites ?

    How is anyone supposed to know the answer to that vague question?

    Tell us what your production environment looks like and the impact of an outage on your business/project and someone may be able to answer you.

    Thanked by 2albertdb GCat
  • exception0x876exception0x876 Member, Host Rep, LIR

    You can use any VM for production websites as long as you have a failover in another dc (and another provider if possible).

    Thanked by 1chrisfe
  • chrisfe said: Should I use them for production websites ?

    Scaleway product line is "production ready",

    beta is finish since April 02 2015 (near 1 year) with no known service outrage on our side :-)

    Thanked by 1chrisfe
  • NekkiNekki Veteran

    exception0x876 said: You can use any VM for production websites as long as you have a failover in another dc (and another provider if possible).

    So production ready = any VM + one other VM with a different provider?

    Thanked by 1support123
  • exception0x876exception0x876 Member, Host Rep, LIR

    @Nekki said:
    So production ready = any VM + one other VM with a different provider?

    I'm ready to be enlightened by your definition of production readiness @Nekki

  • NekkiNekki Veteran

    exception0x876 said: I'm ready to be enlightened by your definition of production readiness @Nekki

    'Production readiness' is a meaningless generalisation and therefore has no definition.

    In your scenario, should VM 'A' become unavailable, you're instantly running without contingency and everything hinges on VM 'B' until you can spin up a new VM and replicate everything across, so you've immediately lost all redundancy.

    Doesn't sound much like any production environment I've ever seen.

  • exception0x876exception0x876 Member, Host Rep, LIR
    edited April 2016

    @Nekki said:
    Doesn't sound much like any production environment I've ever seen.

    It all depends on how fast you can recover old/setup new VM. If that can be done within an hour, 2 VMs failover should be enough. Ofc if we are talking about databases with dozens of terabytes of data, so the recovery takes weeks, 2 VMs might not be enough.

    EDIT: I was talking about an average Joe's scaleway usage scenario.

  • Just keep in mind they dont do backups nor have any RAID installed as far as I am aware..

  • NekkiNekki Veteran

    exception0x876 said: It all depends on how fast you can recover old/setup new VM. If that can be done within an hour, 2 VMs failover should be enough. Ofc if we are talking about databases with dozens of terabytes of data, so the recovery takes weeks, 2 VMs might not be enough.

    Running without redundancy isn't an acceptable situation to many, even if it's only for an hour. You're also assuming that you both know about the issue instantly and are in a position to start resolving it immediately, so I'd say an hour is wholly unrealistic other than in the most simplistic of configurations in the most fortunate scenarios.

    But that aside, your assertion of 'any two VMs' is still inappropriate. Even at a very basic level you should be considering redundancy at the network level and disk configurations.

    I do feel you've missed my point though - there is no 'average joe', and tossing around advice like that is actually harmful to the less knowledgeable.

  • exception0x876exception0x876 Member, Host Rep, LIR

    @Nekki said:
    Running without redundancy isn't an acceptable situation to many, even if it's only for an hour. You're also assuming that you both know about the issue instantly and are in a position to start resolving it immediately, so I'd say an hour is wholly unrealistic other than in the most simplistic of configurations in the most fortunate scenarios.

    You are also assuming that there is no automatic recovery procedure in place.

    @Nekki said:
    But that aside, your assertion of 'any two VMs' is still inappropriate. Even at a very basic level you should be considering redundancy at the network level and disk configurations.

    This is really questionable too, while it is a big plus it is not really a must. Besides most providers do have a redundancy on the network level.

    @Nekki said:
    I do feel you've missed my point though - there is no 'average joe', and tossing around advice like that is actually harmful to the less knowledgeable.

    I'm sorry but it seems your feelings mislead you. I got your point from the start, but it does not mean I agree with it. I believe my advice is useful for most cases and is not harmful even for less knowledgeable. But ofc feel free to disagree with it.

  • trvztrvz Member

    Should I use Scaleway for production websites ?

    No.

  • NekkiNekki Veteran
    edited April 2016

    exception0x876 said: You are also assuming that there is no automatic recovery procedure in place.

    Not an assumption - again turning to your 'any VM' suggestion, lots of providers do not have automatic recovery, so we're looking at the worst-case scenarios here.

    exception0x876 said: This is really questionable too, while it is a big plus it is not really a must. Besides most providers do have a redundancy on the network level.

    It's not 'questionable' in the slightest. The point I was making was that these elements should always be considered for a production environment - of course, not everyone will need redundancy in these areas.

    Thanked by 1MikePT
  • johnmadboyjohnmadboy Member
    edited April 2016

    I think that Scaleway is an excellent hosting for several reasons:

    • It's quite cheap, but it has a good pack;
    • Control panel is easy to use;
    • You can backup servers and backup data through pictures;
    • You can add new SSD;
    • You can create a security groups and the inclusion of servers to administer traffic to the level of its direction and the availability of ports;
    • There is a wide choice of distributions to install and ready server configurations;
    • The possibility of payment server virtual card QIWI, VISA.

    You just need to fill simple forms to create what you wish!

  • @johnmadboy said:
    I think that Scaleway is an excellent hosting for several reasons:

    • It's quite cheap, but it has a good pack;
    • Control panel is easy to use;
    • You can backup servers and backup data through pictures;
    • You can add new SSD;
    • You can create a security groups and the inclusion of servers to administer traffic to the level of its direction and the availability of ports;
    • There is a wide choice of distributions to install and ready server configurations;
    • The possibility of payment server virtual card QIWI, VISA.

    You just need to fill simple forms to create what you wish!

    Reasons not to choose them:

    • it's offline.net
    • no RAID
    • the SSD is network disk
    • almost forgot to mention incompetent offline.net staff
    • premium packet loss



      But hey, if you want PRODUCTION ready servers, don't limit your budget to frigging 3€/m you cheap bastard
  • edited April 2016

    @Mathias said:
    But hey, if you want PRODUCTION ready servers, don't limit your budget to frigging 3€/m you cheap bastard

    There's no need for such rudeness in my opinion. If you are saying this on LOWENDtalk.com, I think you're on the wrong forum.

    OP, if you want production ready stuff, imo, feel free to use Online.net, but make sure you have another provider and find some way to have some form of redundancy (for cheap?)... winks at OVH

    But if it's seriously important stuff and you can't bear to have minimal downtime, I would say, go for the larger providers that are known to have higher uptimes. ahem DO and AWS

    Personally I'm going to work on running two hastebin instances with the same data on 2 NAT VPSes in different DCs, then have at least 2 nodes to reverse proxy them and have the IP changed for the DNS record when the main reverse proxy node goes down.

  • ZweiTigerZweiTiger Member
    edited April 2016

    @Mathias said:

    @johnmadboy said:
    I think that Scaleway is an excellent hosting for several reasons:

    • It's quite cheap, but it has a good pack;
    • Control panel is easy to use;
    • You can backup servers and backup data through pictures;
    • You can add new SSD;
    • You can create a security groups and the inclusion of servers to administer traffic to the level of its direction and the availability of ports;
    • There is a wide choice of distributions to install and ready server configurations;
    • The possibility of payment server virtual card QIWI, VISA.

    You just need to fill simple forms to create what you wish!

    Reasons not to choose them:

    • it's offline.net
    • no RAID
    • the SSD is network disk
    • almost forgot to mention incompetent offline.net staff
    • premium packet loss



      But hey, if you want PRODUCTION ready servers, don't limit your budget to frigging 3€/m you cheap bastard

    Agree. If the ssd fail then say goodbye about your data.

    Better to use 3eur/mo OVH SSD VPS 1 , scaleway is a hobby fail.

    Production ready?

    -Yes it is if you could take all risk about your data... Everybody say then backup backup backup.. but easier and better and time saver if the whole node no fail...

  • daerraghdaerragh Member
    edited April 2016

    I've been using Scaleway for 50 days now, their cheapest VM (VC1S - 2x x86, 2gb, 50gb ssd) as a torrent seedbox. I haven't encountered any downtime/unscheduled reboots since and they advertise the VM as 200mbit guaranteed but I can easily achieve 700mbit throughput.

  • I would never consider a host for production usage that has slow ticket response times or poor support in general.

    If you have contacted their support and feel comfortable with their ability to handle issues, then I would say "yes".

  • @daerragh said:
    I've been using Scaleway for 50 days now, their cheapest VM (VC1S - 2x x86, 2gb, 50gb ssd) as a torrent seedbox. I haven't encountered any downtime/unscheduled reboots since and they advertise the VM as 200mbit guaranteed but I can easily achieve 700mbit throughput.

    This thread is about "production websites", not about running cheap seedboxes.

    If your seedbox data gets lost, you don't care, set it up again and torrent the files again

    So in this case, your testimony doesn't prove anything

    @globalRegisters said:
    I would never consider a host for production usage that has slow ticket response times or poor support in general.

    If you have contacted their support and feel comfortable with their ability to handle issues, then I would say "yes".

    Their ticket response was always quite fast here. But people don't care about ticket response times. You want resolution times. Everyone can hire Indian workforces to reply to all your tickets in less than 5 minutes - 24/7/365. But do they solve problems? no.

    Thanked by 2chrisfe netomx
  • Mathias said: This thread is about "production websites", not about running cheap seedboxes.

    If your seedbox data gets lost, you don't care, set it up again and torrent the files again
    So in this case, your testimony doesn't prove anything

    I merely stated that I'm happy with a Scaleway VM, that it didn't have downtime yet, and that it even exceeds its advertised bandwidth. I don't know if Scaleway is good for production but nevertheless I myself am happy with their offering.

    Thanked by 1chrisfe
  • It would depend what the "Production" workload actually is and how important it is that it's available 24/7.etc

  • TionTion Member

    @globalRegisters said:
    I would never consider a host for production usage that has slow ticket response times or poor support in general.

    Thats the very reason why OVH isn't a bit more production ready than Online.net.

  • You should use both scaleway and ovh for production.

  • cfgguycfgguy Member, Host Rep

    chrisfe said: Should I use them for production websites

    There are various definitions for Production websites.
    1) Like a Forum for fun can also be production website
    2) You earning your living from this site can also be production website.
    3) You and other 10 people are dependent on this production website for their living.

    Find the complexity and importance and judge accordingly. Making profits on lower expensive servers is a good thing but it should not risk the business.

  • I am using c1 for production website since its launch.. but of course it is not of a critical nature and I am having daily backup to remote server.
    I didn't had any major issues till now expect few network problems few months ago with minor downtimes.

  • I can't recommend them anymore. The last three C2 series I have tried to spin up for production based services have all failed to be created and have required support to manually review. They also charge you for the server while they investigate, and if you get rid of it they just give up looking into why the machine couldn't be created.

    I have a ticket that is still open from 3/22 regarding being unable to add additional volumes to an existing production machine. I am migrating away this week.

  • @Mathias said:

    Reasons not to choose them:

    • it's offline.net
    • no RAID
    • the SSD is network disk
    • almost forgot to mention incompetent offline.net staff
    • premium packet loss



      But hey, if you want PRODUCTION ready servers, don't limit your budget to frigging 3€/m you cheap bastard

    Your’s reaction was a little rough, but you're right!

    Thanked by 1Mathias
  • @ZweiTiger said:
    Agree. If the ssd fail then say goodbye about your data.

    Better to use 3eur/mo OVH SSD VPS 1 , scaleway is a hobby fail.

    Production ready?

    -Yes it is if you could take all risk about your data... Everybody say then backup backup backup.. but easier and better and time saver if the whole node no fail...

    Good notice, I didn’t think about it that time.

Sign In or Register to comment.