Skip to main content

I'm trying to run the 'FME Server Workflow Management' workspaces and running into a couple of issues running 'chain_job.fmw' depending on how I reference the server itself in the Web Connections settings.

1. Web Connection configured as same domain I publish against (eg http://site.company.com)

  • running chain_job from desktop and calling the three published workspaces runs fine.
  • running the published chain jobs errors with a web pages content
    • <title>Notification: DNS Failure</title>
    • <h1>This Page Cannot Be Displayed</h1>
    • The host name resolution (DNS lookup) for this host name( site.company.com ) has failed. The Internet address may be misspelled or obsolete,the host ( site.company.com ) may be temporarily unavailable,or the DNS server may be unresponsive.
    • Date: Tue, 20 Feb 2018 16:22:11 AEDT<br />Username: <br />Source IP: 99.99.99.99(this was the machines IP)<br />URL: POST http://site.company.com.com/fmerest/transformations/commands/transact/chaining/load_data.fmw<br />Category: <br />Reason: UNKNOWN<br />Notification: DNS_FAIL
  • The machines appears to see itself ok and can reach the server interface when I'm logged into it remotely (as myself)

2 .Web Connection configured as FME Servers IP address

  • running chain_job from desktop and calling the three published workspaces runs fine.
  • running the published chain jobs errors with
    • INFORM|FMEServerJobSubmitter_1(ServerFactory): http://99.99.99.99 - Running workspace 'load_data.fmw' in repository 'chaining'...
    • WARN |FMEServerJobSubmitter_1(ServerFactory): http://99.99.99.99 - Failed to submit request to run workspace 'load_data.fmw' in repository 'chaining'
    • WARN |FMEServerJobSubmitter_1(ServerFactory): Reason - user 'sjhtest' is not authorized to perform this action

I'm not sure what I'm missing here, in issue one is there somewhere I should be ensuring the dns works? I've tried adding it to the hosts file, which seemed unnecessary as I could see if when logged in.

In issue two could the authentication be failing because it's not on the expected domain? although I accessed the token generation using the IP path and it didn't have any issues.

Thanks for any help!

 

 

 

Hi @stevenjh, this is a very interesting problem (x2)! Are you able to remote into the machine where FME Server resides, and run chain_job.fmw from an FME Desktop installation (local to that same machine)?

 

 

*If so, I would be curious if the response differs if you configure an FME Server Web Connection with localhost, FQDN, or IP address.

 


Hi @stevenjh,

There are a couple of things I might look into.

1) A proxy may be interfering with the DNS. https://docs.safe.com/fme/html/FME_Server_Documentation/Content/AdminGuide/Using_FME_Server_with_Proxy_Server.htm?Highlight=Proxy

2) It may have to do with the accounts running the FME Server. Are they run under localhost? You can try changing the service account. https://docs.safe.com/fme/html/FME_Server_Documentation/Content/AdminGuide/Running_System_Services_Under_Different_Accounts.htm?Highlight=service%20account

If you can provide the log file from FME Server though that may be able to give us an indication on why the DNS is not being resolved.

Thanks,


Hi @stevenjh,

There are a couple of things I might look into.

1) A proxy may be interfering with the DNS. https://docs.safe.com/fme/html/FME_Server_Documentation/Content/AdminGuide/Using_FME_Server_with_Proxy_Server.htm?Highlight=Proxy

2) It may have to do with the accounts running the FME Server. Are they run under localhost? You can try changing the service account. https://docs.safe.com/fme/html/FME_Server_Documentation/Content/AdminGuide/Running_System_Services_Under_Different_Accounts.htm?Highlight=service%20account

If you can provide the log file from FME Server though that may be able to give us an indication on why the DNS is not being resolved.

Thanks,

I was pretty sure we're not running proxies for any of our access so jumped straight to option 2. I set the Engine and the Application Server to run as myself ( just for expediency ) and the connection back to the FQDN, clicked run on Chain jobs and it all worked as expected. Thanks!

 

 


Reply