[ 
https://issues.apache.org/jira/browse/BROOKLYN-588?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16490648#comment-16490648
 ] 

Aled Sage commented on BROOKLYN-588:
------------------------------------

There are two workarounds.

The recommended workaround is to use a more recent version of CentOS. On AWS, a 
good choice is the most recent centos.org image from the AWS marketplace (see 
https://aws.amazon.com/marketplace/pp/B00O7WM7QW). However, this involves first 
subscribing to it in the marketplace. The Amazon Linux AMI is another good 
choice, but this is not a normal CentOS image so it depends what distro(s) the 
entity was developed/tested against.

Another workaround is to change your blueprint to first do {{sudo yum update -y 
curl nss}}, before the {{curl}} command is executed.

> SoftwareProcess download with curl can fail on CentOS 7.0 (TLS negotiation)
> ---------------------------------------------------------------------------
>
>                 Key: BROOKLYN-588
>                 URL: https://issues.apache.org/jira/browse/BROOKLYN-588
>             Project: Brooklyn
>          Issue Type: Bug
>    Affects Versions: 0.12.0
>            Reporter: Aled Sage
>            Priority: Major
>
> When a {{SoftwareProcess}} entity needs to download an install artifact, it 
> often uses curl.
> When running CentOS 7.0, this can fail. For example, when attempting to 
> download something from github:
> {noformat}
> /usr/bin/curl
> curl: (37) Couldn't open file 
> /home/users/amp/.brooklyn/repository/EtcdNode/2.3.1/etcd-v2.3.1-linux-amd64.tar.gz
>   % Total    % Received % Xferd  Average Speed   Time    Time     Time  
> Current
>                                  Dload  Upload   Total   Spent    Left  Speed
>   0     0    0     0    0     0      0      0 --:--:-- --:--:-- --:--:--     0
>   0     0    0     0    0     0      0      0 --:--:-- --:--:-- --:--:--     0
>   0     0    0     0    0     0      0      0 --:--:-- --:--:-- --:--:--     0
> curl: (35) Peer reports incompatible or unsupported protocol version.
>   % Total    % Received % Xferd  Average Speed   Time    Time     Time  
> Current
>                                  Dload  Upload   Total   Spent    Left  Speed
>   0     0    0     0    0     0      0      0 --:--:-- --:--:-- --:--:--     0
>   0     0    0     0    0     0      0      0 --:--:-- --:--:-- --:--:--     0
>   0     0    0     0    0     0      0      0 --:--:-- --:--:-- --:--:--     0
> curl: (22) The requested URL returned error: 404 Not Found
> Could not retrieve etcd-v2.3.1-linux-amd64.tar.gz. Tried: 
> file://$HOME/.brooklyn/repository/EtcdNode/2.3.1/etcd-v2.3.1-linux-amd64.tar.gz,
>  
> https://github.com/coreos/etcd/releases/download/v2.3.1/etcd-v2.3.1-linux-amd64.tar.gz,
>  
> http://downloads.cloudsoftcorp.com/brooklyn/repository/EtcdNode/2.3.1/etcd-v2.3.1-linux-amd64.tar.gz
> Executed 
> /tmp/brooklyn-20180521-195405819-Dfo2-installing_EtcdNodeImpl_id_oe3.sh, 
> result 9
> {noformat}
> This can happen when using a 'minimal' location in AWS (e.g. when just 
> specifying the {{osFamily: centos}}, and not an explicit AMI, which defaults 
> to a CentOS 7.0 AMI).



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

Reply via email to