Skip to main content

Backup Transport can not validate with AWS S3

Nate Reist
Hey there! I'm new here and I apologize if this isn't the correct forum to place this question in, if so let me know where I should post it and I will move it there. I'm having an issue with the cp backup transport not connecting to amazon to store the backups of my server. Up until yesterday this appeared to be working fine, but this morning it stopped moving backup files there and will no longer validate the credentials. I am sure the credentials are still valid as I have another server that still connects and backs up fine using the same credentials. I believe this maybe an SSL issue as I have looked through the logs and found these errors: Upload attempt failed: Amazon::S3: Amazon responded with 500 Can't connect to s3-us-west-1.amazonaws.com:443 at /usr/local/cpanel/3rdparty/perl/526/lib64/perl5/cpanel_lib/Amazon/S3/Bucket.pm line 136. ; And error aborting upload: Amazon::S3: Amazon responded with 500 Can't connect to s3-us-west-1.amazonaws.com:443 at /usr/local/cpanel/3rdparty/perl/526/lib64/perl5/cpanel_lib/Amazon/S3/Bucket.pm line 213.
Path creation failed: Can't connect to s3-us-west-1.amazonaws.com:443 LWP::Protocol::https::Socket: hostname verification failed at /usr/local/cpanel/3rdparty/perl/526/lib64/perl5/cpanel_lib/LWP/Protocol/http.pm line 46. : Can't connect to s3-us-west-1.amazonaws.com:443
Another test using wp-cli brings up another SSL error ( maybe this isn't related by I thought it could be helpful to include the information ) $ wp core verify-checksums Error: Failed to get url 'https://api.wordpress.org/core/checksums/1.0/?version=4.9.7&locale=en_US': cURL error 51: SSL: no alternative certificate subject name matches target host name 'api.wordpress.org'.
I have Root WHM access, here are the server details: /etc/redhat-release:CentOS release 6.10 (Final) /usr/local/cpanel/version:11.72.0.10 /var/cpanel/envtype:virtuozzo
I've verified the hostname for the server has a valid working SSL from cPanel which doesn't expire until November. I updated the Service SSL Certificate after this issue started to see if it would resolve it, and restarted chkservd Any ideas on what I might be missing? Thanks in advance!

Comments

5 comments

Date Votes
  • cPanelMichael
    Hello @Nate Reist, Can you open a
    0
  • Nate Reist
    Hello @cPanelMichael, I have been working on this over the past few days and seem to have uncovered something. In doing traceroutes of amazon aws subdomains and api.wordpress.org it was ending up at our server's IP address, which was odd and seemed like a networking issue. So I reviewed the /etc/hosts and /etc/resolv.conf files, restarted nscd, and tried a couple other things with no success. I did try updating the hostname which had been server.example.com to another sub domain dv02.example.com and the problem seems to have resolved, not sure if server is a subdomain we shouldn't be using. The hosts file still seems to contain part of the old host name (IPs and domains changed): 127.0.0.1 localhost localhost.localdomain localhost4 localhost4.localdomain4 # Auto-generated hostname. Please do not remove this comment. 255.255.255.255 dv02.example.com dv02 server.example.com server example.com example
    We get our WHM / cPanel through our hosting company so I wasn't sure if I could open a support ticket. Can I open a support ticket if we didn't purchase it directly? And should I still do that even though the issue seems to be currently resolved? Thanks.
    0
  • cPanelMichael
    Hello Nate, It's possible that changing the hostname isn't what actually solved the issue. I've seen the "Upload attempt failed: Amazon::S3: Amazon responded with 500 Can't connect" error in the past when there were temporary network issues with Amazon or when they rate limit a specific server. I recommend contacting their support team if the issue reoccurs to see if they have any additional information to offer on why the connection was rejected. Thank you.
    0
  • Nate Reist
    Hey Michael, Thanks for the response. This seems like a weird issue for sure. I could definitely understand the potential of getting blocked by AWS for exceeding rate limits or problems with network, but would you think that api.wordpress.org would be effected at the same time? I guess that could be possible if wordpress.org was using AWS services as well, but it seems coincidental. If it happens again I will contact AWS and see if they can give me any answers. Do you know if requests from the cPanel backup transport actually log anywhere so I could find
    0
  • cPanelMichael
    Hello Nate, POST headers are not logged for the Amazon S3 backup transporter at this time, but we do have internal case CPANEL-13356 open to request the addition of a debug feature that would ensure the POST headers are logged. I don't have a time frame to offer on the potential inclusion of this new feature, but I'll update this thread with more information on the status of this case as it becomes available. Thank you.
    0

Please sign in to leave a comment.

Didn't find what you were looking for?

New post