Who has used Infor Impl. srvcs for HA/Engine upgrades?

This topic has 4 replies, 3 voices, and was last updated 9 years, 11 months ago by Thomas Morris.

Creator

Topic
January 21, 2015 at 6:37 pm #54542
Thomas Morris
Participant
Our Cloverleaf HA environment was established about 5 years ago.

We used Infor Support to help us with the implementation.

We have recently upgraded our Cloverleaf Engine (PROD and TEST) environment(s) from v5.8.7 to v6.0.2.

We DID NOT use Infor Support for this upgrade.

Infor is suggesting that in order ensure the Cloverleaf engine and HA environment(s) will work properly after the engine upgrade, we engage their services for this implementation; they also explain that most of their clients use them to do this. Is this a common experience, or do most organizations do this without engaging Infor?

If so, what are the “gotchas’ beyond making sure both primary and fail-over environments are on the same version?

Any feedback would be greatly appreciated.

#UIH_TAM3
Creator

Topic

Viewing 3 reply threads

Author

Replies
- January 22, 2015 at 3:21 pm #81925
  Steve Williams
  Participant
  Hi Thomas,
  
  I’m just finishing up a site in Chicago-land, upgrading from 5.8.5 to 6.1.0 with 3 IBM HA (fail over) servers in two data centers. We did not engage Infor Svcs but did it “in house.” I’m a contractor, so it wasn’t directly “in house,” but part of my job is to train the FTE’s on failover and update their disaster recovery docs.
  
  There were no direct gotcha issues other than looking for hard coded variables (like /quovadx/cis5.8/… and QUOVADX_INSTALL_DIR) and such in the fail over scripts. I went through every script and rewrote them to work with CL v6.1 and replaced any hard coded directory references with soft coded references using $CL_INSTALL_DIR. We did about a dozen fail over tests and hardware failure simulations before taking our new servers live. The team there now has a good knowledge on the failover process and has a high level of confidence that the system can failover automatically without manual user intervention.
  
  Having said that, there does appear to be a small problem with the IDE app not being able to connect to the fail over server sitting behind an F5 Virtual IP server. I’m not sure it’s a Cl IDE caching issue or an F5 issue, but the work around is to have the IDE connect directly to the fail over server IP address. Once that is done, the IDE can use the Virtual IP as expected.
- January 22, 2015 at 3:36 pm #81926
  Thomas Morris
  Participant
  Hey Steve,
  
  Thank you so much for the information!
- January 22, 2015 at 3:40 pm #81927
  Bob Richardson
  Participant
  Greetings,
  
  Integrator at 5.8.7 soon to be upgraded next month to 6.1.0.
  
  We are now running HA failover on AIX Unix 6 TL9 (two production servers). Hosts are virtualized. We engaged INFOR support for our HA project – paid huge dividends as we tested several scenarios: doing ACTIVE-ACTIVE. Very complex. Also: Global Monitor in the mix – current version 6.1.0. Failover scripts part of the deal.
  
  In the server.ini file we had to add a firewall setting so that at failover time the IDE can connect and find our sites on either virtual host.
  
  The virtual host IP is used. Blanked out our IP but here is the section for your examination:
  
  Note: tunnel_port allows IDE testing tool output to be displayed to us.
  
  [firewall]
  
  rmi_exported_server_port=xxx.xxx.xx.xxx
  
  tunnel_port=14019
  
  INFOR provides a document that discusses and illustrates these firewall settings.
  
  Hope this helps you out.
  
  Enjoy.
- January 22, 2015 at 3:44 pm #81928
  Thomas Morris
  Participant
  Thanks, Bob!
Author

Replies

Viewing 3 reply threads

The forum ‘General’ is closed to new topics and replies.