8 Replies Latest reply on Sep 19, 2018 6:35 AM by Glauber Ribeiro

    Error upgrading 9.2.3 to 10.0.1: coordination service is not available

    Glauber Ribeiro

      I'm having a repeatable problem trying to upgrade a Tableau Server 9.2.3 single server installation to 10.0.1. I consistently get this error, immediately after OK-ing the Tableau Configuration dialogue:

       

      "Worker initialization failed. See the tabadmin.log for details."

       

      Tabadmin.log has: "Error: Coordination Service is not available"

       

      Has anybody here seen this?

       

      Since this is a VM, I can snapshot and go back and try different things. Both these failed the same way:

      • 9.2.3 -> 9.3.0
      • 9.2.3 -> 10.0.0

       

      It feels like something gets messed up in the configuration and zookeeper (the coordination service) is not installed.

       

      What does work is installing 10.0.1 from scratch (rename Tableau directory before installing), then restoring the backup from 9.2.3. That surprised me - I thought restoring the backup would cause the problem to happen.

       

      The only reason this is important to us, is because this is a small system, and the first of four that will be upgraded. I would like to understand the problem before I have, potentially, to deal with it on one of our larger server clusters.

       

      I have contacted Tableau Server, but haven't heard back from them in a couple of days (not unusual), so I thought I'd ask here too.

       

      Any ideas why this happens and how I can prevent it?

       

      Thanks,

       

      glauber

        • 1. Re: Error upgrading 9.2.3 to 10.0.1: coordination service is not available
          Jeff Strauss

          do you do a tabadmin stop prior to trying to upgrade?

          • 2. Re: Error upgrading 9.2.3 to 10.0.1: coordination service is not available
            Glauber Ribeiro

            Not this time. I relied on the uninstall to stop everything.

            I will try that.

            • 3. Re: Error upgrading 9.2.3 to 10.0.1: coordination service is not available
              Xiao-Ping Zhang

              Hello Glauber, I have the same error on Tableau 10.0 server. Could I share your solution for this problem? Thanks.

              • 4. Re: Error upgrading 9.2.3 to 10.0.1: coordination service is not available
                Glauber Ribeiro

                Once it hits, the only solution is do an install from scratch and restore the most recent backup (e.g. the one created during the uninstall.

                 

                I haven’t had this error recently. I’m doing 3 things differently, and I don’t know which one helped (or if it has been coincidence):

                 

                1)  Tabadmin stop, before running the installer.

                 

                2)  Before I do a Tableau upgrade, I log in to each server and make sure that there are exceptions in place on the antivirus, for the folders where Tableau is installed. (In my case, I have to do this every time, because antivirus settings get overridden by company policies.)

                 

                3)  I started running the installer and letting it take care of the uninstall. I used to uninstall manually first, but since more or less recently (v10?) the installer now is able to do the uninstall also.

                 

                Good luck and, if it keeps happening, see if Tableau can assign a technician to do an upgrade with you.

                 

                g

                • 5. Re: Error upgrading 9.2.3 to 10.0.1: coordination service is not available
                  John Kuo

                  Try resetting it by doing:

                   

                  tabadmin stop

                  tabadmin cleanup --reset-coordination

                   

                   

                  2 of 2 people found this helpful
                  • 6. Re: Error upgrading 9.2.3 to 10.0.1: coordination service is not available
                    Glauber Ribeiro

                    This sounds like a good idea for something to add to an upgrade script. I wonder if I should add it to our daily backup/restart script.

                     

                    For the record, for version 10, upgrading without first removing the older version is possible (the installer handles the uninstall), and seems to work better than doing a manual uninstall.

                    • 7. Re: Error upgrading 9.2.3 to 10.0.1: coordination service is not available
                      John Kuo

                      Glauber - do try it and let us know if my suggestion helped.

                       

                      And thanks for the tip about v10 upgrade, I'm on v9.3.7 so an upgrade to v10 is next.

                       

                      Cheers,

                       

                      John

                      • 8. Re: Error upgrading 9.2.3 to 10.0.1: coordination service is not available
                        Glauber Ribeiro

                        Yuk. having similar problem, same server, with 2018.2.1.

                         

                        The REST call to get status says:

                         

                        <?xml version="1.0" encoding="UTF-8" standalone="yes"?>

                        <systeminfo xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">

                            <machines>

                                <machine name="schaistab1">

                                    <repository worker="schaistab1:8060" status="Active" preferred="false"/>

                                    <applicationserver worker="schaistab1:8093" status="Active"/>

                                    <vizqlserver worker="schaistab1:8217" status="Active"/>

                                    <dataserver worker="schaistab1:8421" status="Active"/>

                                    <backgrounder worker="schaistab1:8068" status="Active"/>

                                    <gateway worker="schaistab1:80" status="Active"/>

                                    <hyper worker="schaistab1:8220" status="Active"/>

                                    <searchandbrowse worker="schaistab1:8246" status="Active"/>

                                    <cacheserver worker="schaistab1:8336" status="Active"/>

                                    <filestore worker="schaistab1:8859" status="Active" pendingTransfers="0" failedTransfers="0" syncTimestamp="2018-09-19T13:26:12.284Z"/>

                                    <clustercontroller worker="schaistab1:8077" status="Active"/>

                                    <coordination worker="schaistab1:8230" status="Down"/>

                                </machine>

                            </machines>

                            <service status="Down"/>

                        </systeminfo>

                         

                         

                        tsm status -v:

                         

                        node1: localhost

                                Status: RUNNING

                                'Tableau Server Gateway 0' is running.

                                'Tableau Server Application Server 0' is running.

                                'Tableau Server VizQL Server 0' is running.

                                'Tableau Server Cache Server 0' is running.

                                'Tableau Server Coordination Service 0' is running.

                                'Tableau Server Cluster Controller 0' is running.

                                'Tableau Server Search And Browse 0' is running.

                                'Tableau Server Backgrounder 0' is running.

                                'Tableau Server Data Server 0' is running.

                                'Tableau Server Data Engine 0' is running.

                                'Tableau Server File Store 0' is running.

                                'Tableau Server Repository 0' is running (Active Repository).

                                'Tableau Server Administration Agent 0' is running.

                                'Tableau Server Administration Controller 0' is running.

                                'Tableau Server Service Manager 0' is running.

                                'Tableau Server License Manager 0' is running.

                                'Tableau Server Client File Service 0' is running.

                                'Tableau Server Database Maintenance 0' is stopped.

                                'Tableau Server Backup/Restore 0' is stopped.

                                'Tableau Server Site Import/Export 0' is stopped.

                                'Tableau Server SAML Service 0' is stopped.

                         

                         

                        I did this:

                         

                        PS M:\> tsm stop; tsm topology cleanup-coordination-service

                        Stopping service...

                        Job id is '25', timeout is 30 minutes.

                        Service stopped successfully.

                        Removing non-production Coordination Service ensemble.

                        Job id is '26', timeout is 20 minutes.

                        50% - Validating that there are no pending changes.

                        100% - Removing non-production Coordination Service ensemble id '1'.

                        Finished removing non-production Coordination Service ensemble.

                         

                         

                        And it did not solve the problem. I'll open a  support ticket.