Howdy, Stranger!

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


Is it worth it to pay extra to tech giants for servers?
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.

Is it worth it to pay extra to tech giants for servers?

I have recently used google cloud free tier and downloaded cloud console app in smart phone. When ever i opened it it shows list of current issues, somewhere in some service(like sql). Not my instances.

I used vultr. When ever i open it, also showed current issues or maintenance in dashboard for some region.

So is it worth it to pay extra because all of them promises 99.98% uptime. So with multiple services provided by google point of failure will also increase. And failure in one service may create issue in all services. So overall impact might be higher.

«1

Comments

  • deankdeank Member, Troll

    The major difference is that -

    When you bitch at your host, small(er) hosts will try to actually listen to you.
    Big(ger) hosts will give you canned responses.

    Pay no heed to any claims. One can claim 666% uptime but shit can and will happen. That's how PMS occurs in men.

  • m4num4nu Member, Patron Provider

    One can claim 666% uptime but shit can and will happen.

    See this week's partial AWS outage. Some services were down for hours and they didn't even update the status page. I use their DNS for a few things and can confirm it was down for about 1.5h.

    Things will happen smaller and larger hosts. Difference is that newspapers won't write about the outage on the smaller host.

  • When big host goes down, it is not your problem. When smaller goes out - than it is you to blame for cheapskating. You will not sue the shit out of smaller Joe.

    Thanked by 2lentro devp
  • @Another said:
    I have recently used google cloud free tier and downloaded cloud console app in smart phone. When ever i opened it it shows list of current issues, somewhere in some service(like sql). Not my instances.

    I used vultr. When ever i open it, also showed current issues or maintenance in dashboard for some region.

    So is it worth it to pay extra because all of them promises 99.98% uptime. So with multiple services provided by google point of failure will also increase. And failure in one service may create issue in all services. So overall impact might be higher.

    In my Professional life I had to deal with very different providers.

    Smaller providers tend to care about you as a customer. They might not have the perfect support (in case of response time, so don't expect everything repaired in 15 minutes).

    Some mid-range providers won't care about you as a customer, but they will still deliver really good support.

    Then there is aws, Google, Microsoft, oracle, etc. They don't really want you as a customer. They want big corporations using their "lock-in" services and then they want to charge you 10k USD per month for a service which would costs 5k with a small or mid tier provider. They support is also extremely bad. So don't expect a instant response.

    Thanked by 2ariq01 devp
  • @m4nu said:

    One can claim 666% uptime but shit can and will happen.

    See this week's partial AWS outage. Some services were down for hours and they didn't even update the status page. I use their DNS for a few things and can confirm it was down for about 1.5h.

    Things will happen smaller and larger hosts. Difference is that newspapers won't write about the outage on the smaller host.

    And they lied. They are saying that just us-east was down. Why did I then had issues with eu-central and the middle east?

  • better than cociu.

  • The appeal of using large cloud providers isn't really about reliability, even if that might be the common "business" reason.

    Most companies that use Azure, GCloud, AWS, etc. also use other services like managed k8s, load balancers, CDN services, etc. While you can deploy or purchase these services elsewhere, it's often much easier to just buy from the same company. For larger companies, the "enterprise" features like advanced IAM, compliance standards and access to much more powerful instances at hourly rates is also a huge motivator.

    If you're just running a few services on a VM, there isn't really any reason to go with "cloud". Just shove it onto a VPS and let it be.

  • Depends on your use and budget. I have got good deals here in LET in terms of speed and price than vultr, digital ocean and AWS. Its a Big Club and I ain't in it.

    Thanked by 1devp
  • Better to use many smaller providers across separate data centers and upstream networks. Implement your own scaling/failover.

    Thanked by 4OhJohn Another tux devp
  • LeeLee Veteran

    @Another said: I used vultr. When ever i open it, also showed current issues or maintenance in dashboard

    Having used Vultr for many years I think it's an exaggeration there are is issues every time you open the dashboard. Any large operation will have issues from time to time. DO, Hetzner and Linode I have also used for years and the same applies.

    The wider benefit I find with these providers is that issues are resolved way quicker as it affects a lot more than me so no need to open tickets chasing as I know they are on it.

    The smaller LET type providers I have used in the past are the ones I open tickets, wait days for responses, mixed messages as to what is wrong and so on.

    For those that say "the smaller provider cares about you as a customer", true in some cases but really hit and miss.

  • @randomq said: Better to use many smaller providers across separate data centers and upstream networks. Implement your own scaling/failover.

    This.

    Thanked by 1Another
  • xetsysxetsys Member
    edited December 2021

    I for one am comfortable with my personal data at a bigger host than a smaller one. They also usually have better pricing and offer more value as a whole without compromising on support.

  • AnotherAnother Member
    edited December 2021

    I also feel like using separate data centers and up stream networks should maximize availibility. rather than using separate region from same provider

    Thanked by 1OhJohn
  • lentrolentro Member, Host Rep

    "No one ever got fired for using IBM," is the industry saying, only now it's AWS

  • AWS proves spending more money to a bigger company doesn't directly result in more reliability.

  • jarjar Patron Provider, Top Host, Veteran
    edited December 2021

    Go with larger companies if uptime is truly important.

    You won't get any extra uptime though. But you'll feel like you do because everyone will tell you that you will.

  • DediRockDediRock Member, Patron Provider

    It's really about the response times. Larger hosts have a more robotic response to your needs concerns whereas smaller hosts tend to have a more active communication channel.

  • @m4nu said:

    One can claim 666% uptime but shit can and will happen.

    See this week's partial AWS outage. Some services were down for hours and they didn't even update the status page. I use their DNS for a few things and can confirm it was down for about 1.5h.

    Things will happen smaller and larger hosts. Difference is that newspapers won't write about the outage on the smaller host.

    Seems like no companies actually update those.

    Checking to see if services are down is literally the only thing Twitter is good for

  • @jar said:
    Go with larger companies if uptime is truly important.

    You won't get any extra uptime though. But you'll feel like you do because everyone will tell you that you will.

    Uptime might not be different, but I guess you at least know that Google won't just disappear over night

    Thanked by 2kkrajk devp
  • Daniel15Daniel15 Veteran
    edited December 2021

    @Another said: So is it worth it to pay extra because all of them promises 99.98% uptime

    The promised uptime is usually if you have multiple servers across multiple availability zones, and a way to fail over to hot standby if the main servers go down. If you use a fancy cloud service but only have one server, it's not really advantageous over just a regular VPS. The problem with AWS is that people place all their servers into a single region, and then blame AWS when that region goes down.

    One of the main advantages of "cloud" hosting is that you can scale up instantly on demand, so for example businesses can very quickly add more resources for Black Friday or Christmas or whatever, and then scale back down. My previous employer used Azure autoscaling so that new VMs would automatically spin up when demand goes up, and automatically shut down when demand is low. It was a lot easier than their previous solution which was to install more servers in a physical data center to handle peak load, only to have them idle a lot of the time.

    The thing with AWS is that they have so many different services available under the same brand, which is useful for big customers that need to go through lengthy approvals for every vendor (security review, pricing negotiation, etc). It's a lot easier to have one contract with one provider rather than 10 contracts with 10 different providers.

    Thanked by 3varwww ariq01 devp
  • NeoonNeoon Community Contributor, Veteran

    @deank said:
    The major difference is that -

    When you bitch at your host, small(er) hosts will try to actually listen to you.

    If you bitch, your account will get canned including your services.
    If you ask, you will likely get what you want.

    Thanked by 1devp
  • @deank said:
    The major difference is that -

    When you bitch at your host, small(er) hosts will try to actually listen to you.
    Big(ger) hosts will give you canned responses.

    Pay no heed to any claims. One can claim 666% uptime but shit can and will happen. That's how PMS occurs in men.

    yes.

    hostdoc and hostsolutions are best example.

  • Also be careful of hosts promising super-fast ticket response time. Often they refer to "initial response" which is either automated or L1 agent telling you he is either looking into, or has already forwarded your ticket.

  • Depends on your use case. For security google/aws will beat every other provider however you will pay significantly more and downtime exists everywhere eventually.

  • @lentro said:
    "No one ever got fired for using IBM," is the industry saying, only now it's AWS

    Laughs in former Parler CEO...

    Thanked by 1lentro
  • TimboJonesTimboJones Member
    edited December 2021

    @trycatchthis said:

    @lentro said:
    "No one ever got fired for using IBM," is the industry saying, only now it's AWS

    Laughs in former Parler CEO...

    But you realize AWS gave them chances for months and Parler was seen as either incompetent or outright refusing to cooperate? They were shown ample leniency.

    I'm fairly certain IBM has fired ample clients over the years for failing to comply.

  • jarjar Patron Provider, Top Host, Veteran
    edited December 2021

    @TimboJones said:

    @trycatchthis said:

    @lentro said:
    "No one ever got fired for using IBM," is the industry saying, only now it's AWS

    Laughs in former Parler CEO...

    But you realize AWS gave them chances for months and Parler was seen as either incompetent or outright refusing to cooperate? They were shown ample leniency.

    I'm fairly certain IBM has fired ample clients over the years for failing to comply.

    Everyone held them to different standards. Twitter and Facebook are littered with child porn, racism, all that jazz. They say they try to prevent it or remove it but meh, they never seem to break a sweat for it. Just like the time Cloudflare removed a domain for posts that were first on Facebook and Instagram. People always come up with the justification they need to do what they want. They're usually too busy to talk about how they don't apply the same logic to the others.

    Bottom line is if they don't like you, the reasons will write themselves. If they like you enough, they'll cover your ass. You'll never do enough for someone who has already made their mind about you. There's an algorithm to it that any of us could write out, but it's not worth it. Paying someone who hates you is a bad use of money.

    Anyway, mostly unrelated. Always hated Parler, shit content everywhere. Boomer heaven. Gab is the same. I tried to want to use it I just don't.

    Thanked by 1devp
  • TimboJonesTimboJones Member
    edited December 2021

    @jar said:

    @TimboJones said:

    @trycatchthis said:

    @lentro said:
    "No one ever got fired for using IBM," is the industry saying, only now it's AWS

    Laughs in former Parler CEO...

    But you realize AWS gave them chances for months and Parler was seen as either incompetent or outright refusing to cooperate? They were shown ample leniency.

    I'm fairly certain IBM has fired ample clients over the years for failing to comply.

    Everyone held them to different standards. Twitter and Facebook are littered with child porn, racism, all that jazz. They say they try to prevent it or remove it but meh, they never seem to break a sweat for it. Just like the time Cloudflare removed a domain for posts that were first on Facebook and Instagram. People always come up with the justification they need to do what they want. They're usually too busy to talk about how they don't apply the same logic to the others.

    Bottom line is if they don't like you, the reasons will write themselves. If they like you enough, they'll cover your ass. You'll never do enough for someone who has already made their mind about you. There's an algorithm to it that any of us could write out, but it's not worth it. Paying someone who hates you is a bad use of money.

    Anyway, mostly unrelated. Always hated Parler, shit content everywhere. Boomer heaven. Gab is the same. I tried to want to use it I just don't.

    WHAT!? No, the world is full of fucktards. It's well known that people working on these teams that view the really, really bad content are psychologically fucked with PTSD for life (hence, the need for automation and AI to catch the bad shit). The one story that still sticks with me is that a reviewer tapped out after seeing a video of surgery performed on kids while awake. I never saw it and from time to time when I'm reminded, I'm affected by it.

    Also, what's with showing different content when I reply? Was that an edit between when I loaded the page and hit quote reply? I've seen that happen on LET a few times and not sure which was original and which is the edited update.

    Edit: vanilla should quote reply the contents that was showing when one hits Quote.

    Thanked by 2bulbasaur jar
  • @TimboJones said:
    Also, what's with showing different content when I reply? Was that an edit between when I loaded the page and hit quote reply? I've seen that happen on LET a few times and not sure which was original and which is the edited update.

    Yes, hit me a couple of times. It fetches the current content when you use the Quote link, quite annoying tbh.

    Thanked by 1TimboJones
  • TimboJonesTimboJones Member
    edited December 2021

    @jmgcaguicla said:

    @TimboJones said:
    Also, what's with showing different content when I reply? Was that an edit between when I loaded the page and hit quote reply? I've seen that happen on LET a few times and not sure which was original and which is the edited update.

    Yes, hit me a couple of times. It fetches the current content when you use the Quote link, quite annoying tbh.

    Also caching? After i posted, jar's post was still showing shorter post above mine. After your post, jar's post is now the longer version.

Sign In or Register to comment.