7 Replies Latest reply on Jun 5, 2013 2:43 PM by jdtruman

    What should happen when a 'server' is refreshed?

    jdtruman

      This might be out here already but, I would like to know what others are experiencing when a server refresh occurs when keeping the same IP address, does all the resources refresh correctly, should they?  Thanks, Denise

        • Re: What should happen when a 'server' is refreshed?
          branfarm

          Sounds like you're talking about what happens when you replace old hardware with new hardware, but everything config-wise stays the same.   I've found that on Windows servers the Volumes don't come through very well, and I typically have to add a new volume in and delete the old one.  And as long as the interfaces stay the same name they seem to come across just fine.   With Linux boxes it's a lot easier and I haven't had any problems during a refresh, at least not that I remember.  Admittedly it's been awhile since I've gone through the cycle on a linux machine, but we replaced a Window machine recently and the only issue was the C: volume.

          • Re: What should happen when a 'server' is refreshed?
            Scott Sadlocha

            I have done this with servers as well as network devices. My experience with servers lines up with branfarm's, in that all of the information refreshes upon rediscovery and all goes well, with the exception of volumes. Most of the time I have to delete old volumes and add the new volumes back in through List Resources.

            • Re: What should happen when a 'server' is refreshed?
              jdtruman

              Thanks to both of you, I'm seeing the same. I delete the old drive(s) and sometimes the NIC, then add the resources back after rediscovery.  How does your team handle this process ongoing?  We follow a change process for handling p>v, v>v, new builds, and so on.  Do you use auto discovery on specific subnets, reporting on node changes or a combination of things so you don't miss the refreshes.  I'm in preliminary stages of using the SDK to add and delete nodes, so just curious. 

                • Re: What should happen when a 'server' is refreshed?
                  branfarm

                  We wrap that up into the workflow for replacing the server.  The last step on the checklist is "Verify all interfaces, volumes, and custom pollers are up to date and polling in NPM"

                  • Re: What should happen when a 'server' is refreshed?
                    Scott Sadlocha

                    At my previous company we had a defined Server Commission process (as well as a Server Decommission process). The Server Commission process was tracked through a ticket in Altiris HelpDesk, and was assigned to all pertinent parties for their step. These steps were defined in a Server Commission workflow document that could be used as reference alongside the ticket, and also accompanied by a Readiness Assessment to make sure the server was adequate for its intended need. One step in the process was to assign the ticket to the Solarwinds admin (me) to get the server added to monitoring. This same process was used for server "refreshes", and went a long way towards making sure servers didn't slip through the cracks. Also, a change control process was followed that tied into everything else. It sounds like a lot, but it was all linked together fairly well and if the process was used, servers would rarely get missed. Still, I had scheduled discoveries running to pick up anything that might be missed in rare cases. and I also had a habit of scrolling through groups and the admin console looking for anything out of the ordinary. As time went on, I started to create reports and query the DB to find anything out of the ordinary (volumes or interfaces not reporting, etc.).

                     

                    Hope the information helps.

                  • Re: What should happen when a 'server' is refreshed?
                    jdtruman

                    Great, I'm following similar process using Remedy.  Thanks again..