7 ms·

I share the same experience and opinion. I cannot imagine what a significant argument would be to use AWS instead of GCP that outweighs the benefits and the see

by s-xyz 3y ago

I share the same experience and opinion. I cannot imagine what a significant argument would be to use AWS instead of GCP that outweighs the benefits and the seemingly integrated services. GCP has so far covered 99% of my use cases, and I can testify that I have a lot of advanced cases linked to data management, information security, networking and more. I am open and do not judge AWS, and perhaps its also a matter of what you are used to. However my style and approach to problems is more in alignment with GCP. I am for example of the opinion that the cloud provider should take care of everything linked to infrastructure, maintenance and as well scaling. And all of that should be as simple as possible. There is no need to overcomplicate. In this way I can focus on what makes my product unique, rather than spending my time on over-engineering repetitive features with zero value added.

HideousKojima 3y ago
> cannot imagine what a significant argument would be to use AWS instead of GCP that outweighs the benefits and the seemingly integrated services. Confidence that it will continue to exist in the next 3-10 years?
- Eridrus 3y ago
  What odds would you like to bet that GCP is going to be gone in 10 years?
  - perelin 3y ago
    The sentiment is probably not abou GCP as a whole, but about its trigger happiness when it comes to deprecating services. Fun read: https://steve-yegge.medium.com/dear-google-cloud-your-deprecation-policy-is-killing-you-ee7525dc05dc https://steve-yegge.medium.com/dear-google-cloud-your-deprec...
    - Eridrus 3y ago
      I'm not going to read yegge's rants. Name the odds and condition you think gcp will violate.
      - cachesking 3y ago
        Google IoT came and went.
  - judge2020 3y ago
    If they do want to kill it, I can see them giving as little as a 5 year window before everything is shut down. Odds? Nobody's a fortune teller, but it was still being subsidized by ads with 3B loss in its FY22[0 p. 70], only increasing 100M in profit despite cloud revenue growing from 19B to 26B overall [0 p. 66]. They're either going to need to increase costs or add more "killer" products with large margins if they want to become profitable. 0: https://abc.xyz/assets/d4/4f/a48b94d548d0b2fdc029a95e8c63/2022-alphabet-annual-report.pdf#page=70 https://abc.xyz/assets/d4/4f/a48b94d548d0b2fdc029a95e8c63/20...
    - Eridrus 3y ago
      At what odds are you willing to make a bet on this? Do you think there's a 20% chance it's gone in 5 years? Do you think it's 5%? Do you think it's 30%? People keep saying things will be shut down and I think they should put their money where their mouth is.
      - seanthemon 3y ago
        What an oddly aggressive comment - perhaps the problem isn't prediction but extrapolation of past experiences with Google, would you not agree that it gives you a sliver of worry that they may cut the service off? You seem very intense about Google's offering, maybe you feel attacked by the points made in the thread, but the way you're approaching this discussion is very strange.
        gkbrk 3y ago
        Eridrus used to work at Google. Not sure if it would be good to disclose before aggressively asking people to bet on the lifespan of Google products, I certainly would err on the side of disclosing.
        Eridrus 3y ago
        Sure, I used to work at Google, I used to work at Microsoft, some of my best friends worked at Amazon, I own stock in all of them. Are you interested in making a bet on the chance GCP will be discontinued?
  - HideousKojima 3y ago
    If it can't get profitable (or at least within $1B/year of profitable, or have some sort of major loss leader value to Google) then I'd give it 90% likelihood they they announce a shutdown within 10 years. Whether or not they'll continue to lose billions a year for Google I can't say.
    - Eridrus 3y ago
      If you want to give me 10:1 odds GCP is not going to be around in 10 years, I will bet you any amount of money you are willing to name lol.
    - VirusNewbie 3y ago
      They had a net profit of 266M this quarter, so like, as of today they're on track to go past 1B in profit.
    - elijaht 3y ago
      Cloud was profitable during the last quarter
nerdjon 3y ago
For a fairly simple setup AWS automates a lot of things. If I just want an EC2 instance and a database, the VPC already exists and there is now even a button on an RDS that says something like "connect to ec2" which sets up the security groups on both for you. However the argument against using GCP has already been stated in this thread multiple times. - Will it exist in 3-5 years. Google has a reputation for a reason. - Will the cost increase in the next year. Google (including google cloud) has a reputation for a reason. - Here we have seen time and time again problems reaching support with Google Cloud. Meanwhile my personal AWS account that I spend maybe $5 a month on... I emailed support about a billing problem and got a response about a day later. I would not risk my job by using Google Cloud and being at the whims of Google being Google. The last thing I want is to have to go to my leadership and explain why the cost is going up or why suddenly we need to spend a bunch of engineering effort because google is discontinuing something. If I really didn't want to support AWS/Amazon (which is fair) I would go Azure long before going GCloud. Hell I might be convinced to go IBM but thankfully I don't have to make that decision with Azure.
- kstrauser 3y ago
  Nailed it. When we've had problems with AWS, I could drop our rep an email and be on a conference call with the engineers who run the service a day later. That not for a giant account, either. They'd also periodically reach out to offer free consulting services that would cut our monthly bill. They were upfront about the motivation: they wanted us to integrate more of their services. The upside for us was that we could get better, cheaper service. If you're planning on sticking with them anyway, that's a very attractive offer. Meanwhile, the support venue for our enterprise Google Workspace account was a mutual support web forum as far as I could tell. Unless you've experienced both, it's hard to believe how radically differently the 2 companies handle customer support.
  - cassianoleal 3y ago
    Interesting. A previous client was all-in on GCP. They were not a huge operation and in fact were slowly migrating on-prem stuff to it. GCP support was pretty good. The account manager was always on point and whenever we needed we could have an actual engineer or two on a call with us to troubleshoot stuff and give us advice and general support. Of course, for less urgent and less severe issues it would generally be done over the course of a couple days to a week in support ticket messages / email but that was perfectly fine. I've been back to using AWS for a little while now and I miss how much simpler it is to set up things in GCP. Even IAM which I remember bitching about so much during those days, now I just miss how much simpler it was and I didn't realise. Also if you need Kubernetes, no other offering compares to GKE.
acdha 3y ago
> However my style and approach to problems is more in alignment with GCP. I am for example of the opinion that the cloud provider should take care of everything linked to infrastructure, maintenance and as well scaling. You get that on AWS, too. The difference in my experience has been that once you’re out of a narrow range of features you’re more likely to hit something which is builtin on AWS but has a GCP ticket from 5 years ago with no progress. If you have any kind of security compliance requirements that gets brutal on GCP since you have to take on running an entire service to get that one missing feature.
unethical_ban 3y ago
I used to work in cloud security, mostly AWS but got to hear about experiences with GCP. AWS IAM is much better than GCP, or was 3 years ago. If you are in an industry where fine-tuned RBAC per service is a thing you want, it's better to go AWS, all things equal. Also, GCP reps suck compared to AWS in terms of responsiveness and general give-a-shits. Hearsay.
- _cenw 3y ago
  Yeah, no. GCP actually allows you to give developers full permissions on their project. This sort of separation doesn't exist on AWS, at least not without making it very hard to impossible to share resources. So on AWS, you either have a really good platform team that handles permissions / terraform review / all infra setup (unlikely), a bad or understaffed team doing the same so everyone waits a week or two on specific things being set up (and everyone will hate them for killing their velocity), or you let everyone run wild with way too many permissions so they can actually use the console, hoping they'll stick to best practices (they wouldn't). At least developers doing unsafe things is contained in GCP. But all that aside, AWS is better. It's much more predictable which things will work together from the documentation, the support is better and the IAM is better on paper, if you actually have that unicorn platform/infrastructure team.
  - unethical_ban 3y ago
    Well, I was on a unicorn team, then. Guardrails on all AWS resources, developers deployed via gitlab/terraform to AWS accounts mapped to departments. You can share resources across AWS accounts. I don't know for easy that is on gcp, but it didn't seem crazy on Amazon.
zepolen 3y ago
GCP support is a joke compared to AWS.
- pirsquare 3y ago
  Yep shutdown our production servers and take 1 whole week to get it resolved. Nothing fraudulent on our end. Just missing out on KYC form.
marcinzm 3y ago
GCP is overall a good product. Support is so atrocious and they break things often enough that it more than offsets that in any larger scale use case in my experience.

by s-xyz 3y ago

HideousKojima 3y ago

> cannot imagine what a significant argument would be to use AWS instead of GCP that outweighs the benefits and the seemingly integrated services. Confidence that it will continue to exist in the next 3-10 years?

Eridrus 3y ago
What odds would you like to bet that GCP is going to be gone in 10 years?
- perelin 3y ago
  The sentiment is probably not abou GCP as a whole, but about its trigger happiness when it comes to deprecating services. Fun read: https://steve-yegge.medium.com/dear-google-cloud-your-deprecation-policy-is-killing-you-ee7525dc05dc https://steve-yegge.medium.com/dear-google-cloud-your-deprec...
  - Eridrus 3y ago
    I'm not going to read yegge's rants. Name the odds and condition you think gcp will violate.
    - cachesking 3y ago
      Google IoT came and went.
- judge2020 3y ago
  If they do want to kill it, I can see them giving as little as a 5 year window before everything is shut down. Odds? Nobody's a fortune teller, but it was still being subsidized by ads with 3B loss in its FY22[0 p. 70], only increasing 100M in profit despite cloud revenue growing from 19B to 26B overall [0 p. 66]. They're either going to need to increase costs or add more "killer" products with large margins if they want to become profitable. 0: https://abc.xyz/assets/d4/4f/a48b94d548d0b2fdc029a95e8c63/2022-alphabet-annual-report.pdf#page=70 https://abc.xyz/assets/d4/4f/a48b94d548d0b2fdc029a95e8c63/20...
  - Eridrus 3y ago
    At what odds are you willing to make a bet on this? Do you think there's a 20% chance it's gone in 5 years? Do you think it's 5%? Do you think it's 30%? People keep saying things will be shut down and I think they should put their money where their mouth is.
    - seanthemon 3y ago
      What an oddly aggressive comment - perhaps the problem isn't prediction but extrapolation of past experiences with Google, would you not agree that it gives you a sliver of worry that they may cut the service off? You seem very intense about Google's offering, maybe you feel attacked by the points made in the thread, but the way you're approaching this discussion is very strange.
      - gkbrk 3y ago
        Eridrus used to work at Google. Not sure if it would be good to disclose before aggressively asking people to bet on the lifespan of Google products, I certainly would err on the side of disclosing.
        Eridrus 3y ago
        Sure, I used to work at Google, I used to work at Microsoft, some of my best friends worked at Amazon, I own stock in all of them. Are you interested in making a bet on the chance GCP will be discontinued?
- HideousKojima 3y ago
  If it can't get profitable (or at least within $1B/year of profitable, or have some sort of major loss leader value to Google) then I'd give it 90% likelihood they they announce a shutdown within 10 years. Whether or not they'll continue to lose billions a year for Google I can't say.
  - Eridrus 3y ago
    If you want to give me 10:1 odds GCP is not going to be around in 10 years, I will bet you any amount of money you are willing to name lol.
  - VirusNewbie 3y ago
    They had a net profit of 266M this quarter, so like, as of today they're on track to go past 1B in profit.
  - elijaht 3y ago
    Cloud was profitable during the last quarter

nerdjon 3y ago

For a fairly simple setup AWS automates a lot of things. If I just want an EC2 instance and a database, the VPC already exists and there is now even a button on an RDS that says something like "connect to ec2" which sets up the security groups on both for you. However the argument against using GCP has already been stated in this thread multiple times. - Will it exist in 3-5 years. Google has a reputation for a reason. - Will the cost increase in the next year. Google (including google cloud) has a reputation for a reason. - Here we have seen time and time again problems reaching support with Google Cloud. Meanwhile my personal AWS account that I spend maybe $5 a month on... I emailed support about a billing problem and got a response about a day later. I would not risk my job by using Google Cloud and being at the whims of Google being Google. The last thing I want is to have to go to my leadership and explain why the cost is going up or why suddenly we need to spend a bunch of engineering effort because google is discontinuing something. If I really didn't want to support AWS/Amazon (which is fair) I would go Azure long before going GCloud. Hell I might be convinced to go IBM but thankfully I don't have to make that decision with Azure.

kstrauser 3y ago
Nailed it. When we've had problems with AWS, I could drop our rep an email and be on a conference call with the engineers who run the service a day later. That not for a giant account, either. They'd also periodically reach out to offer free consulting services that would cut our monthly bill. They were upfront about the motivation: they wanted us to integrate more of their services. The upside for us was that we could get better, cheaper service. If you're planning on sticking with them anyway, that's a very attractive offer. Meanwhile, the support venue for our enterprise Google Workspace account was a mutual support web forum as far as I could tell. Unless you've experienced both, it's hard to believe how radically differently the 2 companies handle customer support.
- cassianoleal 3y ago
  Interesting. A previous client was all-in on GCP. They were not a huge operation and in fact were slowly migrating on-prem stuff to it. GCP support was pretty good. The account manager was always on point and whenever we needed we could have an actual engineer or two on a call with us to troubleshoot stuff and give us advice and general support. Of course, for less urgent and less severe issues it would generally be done over the course of a couple days to a week in support ticket messages / email but that was perfectly fine. I've been back to using AWS for a little while now and I miss how much simpler it is to set up things in GCP. Even IAM which I remember bitching about so much during those days, now I just miss how much simpler it was and I didn't realise. Also if you need Kubernetes, no other offering compares to GKE.

acdha 3y ago

> However my style and approach to problems is more in alignment with GCP. I am for example of the opinion that the cloud provider should take care of everything linked to infrastructure, maintenance and as well scaling. You get that on AWS, too. The difference in my experience has been that once you’re out of a narrow range of features you’re more likely to hit something which is builtin on AWS but has a GCP ticket from 5 years ago with no progress. If you have any kind of security compliance requirements that gets brutal on GCP since you have to take on running an entire service to get that one missing feature.

unethical_ban 3y ago

I used to work in cloud security, mostly AWS but got to hear about experiences with GCP. AWS IAM is much better than GCP, or was 3 years ago. If you are in an industry where fine-tuned RBAC per service is a thing you want, it's better to go AWS, all things equal. Also, GCP reps suck compared to AWS in terms of responsiveness and general give-a-shits. Hearsay.

_cenw 3y ago
Yeah, no. GCP actually allows you to give developers full permissions on their project. This sort of separation doesn't exist on AWS, at least not without making it very hard to impossible to share resources. So on AWS, you either have a really good platform team that handles permissions / terraform review / all infra setup (unlikely), a bad or understaffed team doing the same so everyone waits a week or two on specific things being set up (and everyone will hate them for killing their velocity), or you let everyone run wild with way too many permissions so they can actually use the console, hoping they'll stick to best practices (they wouldn't). At least developers doing unsafe things is contained in GCP. But all that aside, AWS is better. It's much more predictable which things will work together from the documentation, the support is better and the IAM is better on paper, if you actually have that unicorn platform/infrastructure team.
- unethical_ban 3y ago
  Well, I was on a unicorn team, then. Guardrails on all AWS resources, developers deployed via gitlab/terraform to AWS accounts mapped to departments. You can share resources across AWS accounts. I don't know for easy that is on gcp, but it didn't seem crazy on Amazon.

zepolen 3y ago

GCP support is a joke compared to AWS.

pirsquare 3y ago
Yep shutdown our production servers and take 1 whole week to get it resolved. Nothing fraudulent on our end. Just missing out on KYC form.

marcinzm 3y ago

GCP is overall a good product. Support is so atrocious and they break things often enough that it more than offsets that in any larger scale use case in my experience.