Using the new 8x pipelines give me HTTP errors

[Kikadass]

Recently there was a new announcement on Atlassian to allow 4x and 8x pipelines in their servers with 16CPUs and 32GB of memory: 

https://bitbucket.org/blog/announcing-our-new-ci-cd-runtime-with-up-to-8x-faster-builds

I was trying to update my pipelines and I also updated the Corporate firewall to accept the new IPs listed here: 

https://support.atlassian.com/bitbucket-cloud/docs/what-are-the-bitbucket-cloud-ip-addresses-i-should-use-to-configure-my-corporate-firewall/

But I still get HTTP issues when trying to download my Bitbucket repository:

This only happens when I use the new 8x or 4x pipelines, if I change the size to 2x it works. 

I also gave more memory to the docker image up to memory: 27128

Any advice on where should I start looking? Is this a known issue?

[Ben · Accepted solution]

Hi Kikadass, 

The documentation is in the process of being updated as it is not made exactly clear how to ensure that your 4x/8x pipeline builds use only the IP addresses within the atlassian-ip-ranges list that you've mentioned.

To do so and avoid firewall issues - you will need to modify your YML configuration as per this article as follows:

Example - using the runtime globally to enable atlassian-ip-ranges for all 4x, 8x steps

options:
  runtime:
    cloud:
      atlassian-ip-ranges: true

Example - using the runtime to enable atlassian-ip-ranges for a particular step

pipelines:
  default:
    - step:
        size: 4x
        runtime:
          cloud:
            atlassian-ip-ranges: true
        script:
          - echo "I use atlassian-ip-ranges"

Hope this helps.

Cheers!

- Ben (Bitbucket Cloud Support)

[Kikadass]

Thank you Ben! 
I think that works. 

Also out of curiosity, what is the max I can set the docker image memory to be? 
I set it to 27128 as I saw examples using that much. 
However, the new 8x size runners have 32GB right? Can I set the docker container to use all that? Or do I have to leave some free for the runner? 

Thank you for your support!
Enrique Piera Serra

[Ben]

Hi @Kikadass 

The 4x and 8x pipelines have the following capabilities:

• 4x builds will be provisioned with 8 CPU cores and 16GB of memory.
• 8x builds will be provisioned with 16 CPU cores and 32GB of memory.

You'll definitely need to leave some room for the build container, by default this uses 1024MB - so the service container could have a maximum of 31744MB assigned.

You may experience build memory limit timeouts, however, so I would suggest experimenting a little bit with this. More information regarding service container memory limits can be found below:

• https://support.atlassian.com/bitbucket-cloud/docs/databases-and-service-containers/

Let me know if you need further clarification on this and I'll be happy to assist! :)

Cheers!

- Ben (Bitbucket Cloud Support)

[Kikadass]

Thank you for all the information, Ben. 

Have a good day!

[Kikadass]

Hello Ben, 
I am not sure whether to ask this here or make a new Question. Please let me know. 

My Bitbucket Cloud hosted runners work fine now, thanks to the information you shared. 

However, the same pipeline on Self hosted runners (also on 8x and memory: 27128) when compiling C++ code with make, the more CPUs I try to use the slower it gets:

make -j4 -> 6min
make -j8 -> 20min
make -j15 -> 31min

This does not happen on developers's machines, only on the Self hosted runner.

Would you know anything about this?

Thanks a lot, 
Enrique Piera Serra

[Ben]

Hi @Kikadass 

I can't see a Bitbucket Cloud account associated with the email address you used to log in and raise this question.

Given you are using the 4x/8x pipelines feature, I am assuming that you either pay for or belong to a workspace on a Standard/Premium plan. If this is the case, I'd highly suggest raising a ticket with our support team directly using this paid workspaceID as any issues such as this will require investigation/potential escalation by our engineering team:

• https://support.atlassian.com/contact/

If you are having issues raising a ticket, please reply to this thread by logging in using the email address belonging to your Bitbucket Cloud account, and let me know your timezone so I can raise it on your behalf with the appropriate team in your geo.

Cheers!

- Ben (Bitbucket Cloud Support)

[enrique piera serra]

Hello Ben, 

I am the same Kikadass :D
I am in South Korea, so timezone would be KST.

Would you be able to help me raise the ticket? 
What information would you need from me?

Thank you for your help! :)

Enrique 

[Ben]

Hi @enrique piera serra !

You're almost in the same timezone as me (GMT+10) so I've raised a ticket on your behalf and assigned it to myself - we will continue our conversation there :)

Cheers!

- Ben (Bitbucket Cloud Support)