14 Replies Latest reply on Nov 5, 2013 6:50 AM by Benjamin THIEN

    Error: retrying 40 of 40 with PXE


      Hello everyone,


      For months we have been plagued by issues with provisioning. Using PXE, we boot into WinPE, enter our credentials, select the desired template, it tries 40 times, times out, and asks you to select a template again. Sometimes it works after a few tries, sometimes it takes hours to start. We recently upgraded to Landesk 9 and are still seeing the issue and I'm hoping someone can shed some light on what our problem may be.


      I've linked and attached 3 ldprovision.log files in case they provide any helpful information. One is from our 8.8 SP4 core and 2 from our new 9 SP2 core.







        • 1. Re: Error: retrying 40 of 40 with PXE
          ThomasCollignon Apprentice


          What is you Database engine (SQL /Oracle)?

          Memory allowed for SQL /Oracle database? (LANDesk prerequisites is 2Mo /node.)

          Do you have a limitation in the number of connections to the database?


          How many PXE Representative is installed? Try with only one.


          If you schedule the template provisioning before connecting the machine to PXE, the template is automatically launched it under WinPE?


          Do you select "Locked Template" or not?




          • 2. Re: Error: retrying 40 of 40 with PXE

            Thanks for the quick response, Unfortunately, I'm not the server administrator but I've asked him for answers to your Oracle/SQL questions. I should have responses in a day or 2.


            As for your other questions, we do have quite a few PXE reps in our environment (currently 8) because we currenlty need 1 PXE rep per subnet. We do only have 1 rep per subnet.


            At the template selection screen we are selecting the templates which are not locked (which generates a locked copy with a timestamp)


            I'll try scheduling the template and get back to you tomorrow.


            Thanks for your help,



            • 3. Re: Error: retrying 40 of 40 with PXE

              Here are his responses:


              >>What is you Database engine (SQL /Oracle)?

              SQL Server 2008 R2


              >>Memory allowed for SQL /Oracle database? (LANDesk prerequisites is 2Mo /node.)

              Set at default (2147483647)


              >>Do you have a limitation in the number of connections to the database?

              Set at default (0)

              • 4. Re: Error: retrying 40 of 40 with PXE
                ThomasCollignon Apprentice

                Ok, nothing special for SQL.
                By default, the Oracle database are created with very few resources...


                Can you have make a test with Scheduled Template?

                SQL Server is on Core Server? If yes LANDesk prerequisite 4Go for System and Landesk process + 2Mb / nodes for memory (if you have 3000 nodes, you need 10Gb of memory). If you have more than 4000 nodes, LANDesk recommends 1 server for LANDesk Core Server, and 1 server for SQL Database.


                Try also this solution:

                Uninstall all your PXE representative, and make a test with only one PXE representative.


                Extracted from LANDesk article:

                To many PXE Reps to update with the Provisioning information for the device
                • If there are a large number of PXE representatives in the environment, this error will occur as the core tries to update them all with the needed information. If it cannot complete in time for the 40 retries, the task will not start. However, if you keep waiting the template may start.


                • Reduce the number of PXE reps and remove all PXE Reps that are not active.


                Each time that I had this problem at my customers, it was one of those causes.


                I hope this can help you!!!

                • 5. Re: Error: retrying 40 of 40 with PXE

                  Thanks again for your response. I'm looking into see if we need to allocate more resources to our Core and/or SQL server.


                  Unfortunately, I'm with just one of the departments  who uses PXE at our university so I can't remove all other reps. I have  removed our non-essential reps for testing so we are now down to 3 PXE  reps instead of 7. I'll see if that helps.


                  I'll report back tomorrow with the more answers and the results of testing with fewer pxe reps.


                  In your experience, how many PXE reps is too many or is it dependent on the resources allocated to the Core? We are down to 6 reps and I can't remove any more.

                  • 6. Re: Error: retrying 40 of 40 with PXE
                    ThomasCollignon Apprentice

                    Hi Georges,

                    You are welcome, if I can help...

                    I know it's difficult, especially when we are not the product administrator...


                    It's just for test... To find where is your problem...

                    Try with 3 PXE repository, but make sure that all are strated (computers and PXE service).

                    Difficulties can appear if some PXE representative are not started...

                    You can also try a very simple template with a only one action in the phase "WinPE".


                    If with Scheduled task all work fine, it's generally related to the difficulty of the Web service to create the actions for the selected template.



                    How many nodes in database?

                    SQL Server 2008R2 and LDMS Core server are installed on the same server?

                    How many CPU?

                    How many memory?

                    Network speed SQL Server / LDMS Core?




                    • 7. Re: Error: retrying 40 of 40 with PXE

                      We had the exact same issues, although we did not contact LANDesk Support yet.  Here is what we did and it finally started working: under the Logs folder on Core server, under Provisioning folder, make sure LogLevel ini file is set to


                      Also delete the "provisioning" log file that is in there.  In LANDesk search for the MAC addresses of the machines that you are trying to provision, and delete them out.  May need to change the scope to include machines that do not have an OS listed, several times it listed "ghost" (minix989234823) machines with no other data.

                      • 8. Re: Error: retrying 40 of 40 with PXE

                        Another update: Regarding resources it looks like we should have enough. We have just over 2000 nodes.


                        Network shoud be plenty fast as well. We have a mix of gigabit and 100mbit connections accross campus.


                        >>Is the SQL Server on the Core Server?

                        No, each instance is running on its own VM.


                        >>What resources (CPU, RAM and Hard drives) have been allocated to the Core Server and SQL Server)?

                        Core – Xeon E5310 1.6 GHz (4 processors), 8 GB RAM, 40 GB (OS), 150 GB (Data)

                        SQL – Xeon E5504 2.0 GHz (4 processors), 16 GB RAM, 40 GB (OS), 350 GB (Data)




                        Just some more information to add:


                        Boot media (CDs and flash drives) are timing out in WinPE as well, not just when booted via PXE.

                        • 9. Re: Error: retrying 40 of 40 with PXE
                          ThomasCollignon Apprentice

                          Ok Georges,


                          LDMS Architecture seems to be correct.

                          It's verry strange...


                          It's logical that the boot on UCB or CD has the same result as the PXE boot.

                          When you select a Provisioning Themple, a Scheduled task is created after a locked copy of the Template has been created.


                          But you have not tested with a scheduled task...

                          • Open LANDesk consol,
                          • Open Provisioning Tool,
                          • Right click on the Provisioning Template you want to play (a scheduled task is created...),
                          • Assign the device you want to apply the template at the task,
                          • Change the Delivery methode if necessary,
                          • Click on Start now, and wait that the policy task is
                          • Wait that the policy is published and it spends waiting to be processed,
                          • Start the targeted computer,
                          • Boot on PXE,
                          • Select "Win PE provisioning",
                          • ...and look that happens...


                          If you want to apply the template to a device that is not yet in the database, two solutions:

                          1. Create a new "Bare Metal" entry:
                            • In LANDesk Console, select "Configuration" in "Network View" (not in the menu, where is located "Devices" and "Queries"). if you have not acces to "Configuration", forgot this solution.,
                            • Right click on "Bare Metal Server",
                            • Create a new entry,
                            • Give the Computer Name,
                            • Give the MAC Address identifier,
                            • Add this new entry,
                            • Save, and schedule on this new node.
                          2. PXE entry:
                          • Search in the list of "All Devices" the entry corresponding to your MAC Address.
                          • If you can not find it, the problem is here ...

                          As pointed out "fxsti01", you must have a Scope for your LANDesk profile which does not allow you to see the PXE entries.
                          Ask your administrator LANDesk to expand your scope accordingly.


                          You can also search your scheduled tasks spots provisioning, and watch the target assigned to this task. If no target appears, confirm that your Scope does not allow you to see these posts (and therefore the task of planning ...)


                          I hope that my translation is clear and correct, and can help you.



                          • 10. Re: Error: retrying 40 of 40 with PXE

                            I can confirm that scheduling does work. I successfully scheduled 4 provisioning templates. I used the PXE entry to schedule and it works without the timeout errors.

                            • 11. Re: Error: retrying 40 of 40 with PXE

                              We just introduced Provisioning in our LD 9.0 SP2 environment and are experiencing the same issue.



                              VERBOSE             TemplateFinder                                7/5/2011 2:49:12 PM       : >>GetTemplateForServer,computerIDN 5654

                              DEBUGGING      TemplateFinder                                7/5/2011 2:49:12 PM       : >>GetProvisioningTaskForComputer.computerIdn 5654

                              INFO      TemplateFinder                                7/5/2011 2:49:12 PM       : No task found. Check task with Done status.

                              ERROR  TemplateFinder                                7/5/2011 2:49:12 PM       : couldn't find task for computer 5654

                              ERROR  ProvisioningService                         7/5/2011 2:49:12 PM       : Unable to find template for computer IDN 5654

                              • 12. Re: Error: retrying 40 of 40 with PXE
                                slingblade Rookie

                                I just went through a week of troubleshooting for the same problem. We also boot to WinPE, enter credentials and then select a provisioning template to run. The retry problem seemed to be inconsistent until we started to think about the problem more critically.


                                The post by fxsti01 hits the problem right on the head. Our techs only had access to our "Workstations" scope. The machines that were being provisioned did not have an existing record in LD. As a result, when provisioning was attempted and the miniscan ran; the devices were outside of the Workstations scope. To solve this problem, we created an additional scope called "Unassigned Devices" and gave access to our techs. This scope is defined as:


                                "Computer"."Device ID" = "Unassigned"  AND  "Computer"."Scan Type" NOT LIKE "Unmanaged Device"


                                (The reason that we excluded Unmanaged Device scans is that we are currently using that to discover/deploy server agents.)


                                I stumbled on the post by fxsti01 just hours after we figured this out and I want to confirm that this was our problem! Hopefully this will help someone else.

                                • 13. Re: Error: retrying 40 of 40 with PXE
                                  Benjamin THIEN Specialist



                                  I already have the same problem, for me i delete a source program use in provisioning and it failed top find template, do you use software distribution in your provisioning ?

                                  Check if all your sources are still ok, try to duplicate too your template.

                                  This is very simple check but for me it works, it was a very little thing to fix...


                                  • 14. Re: Error: retrying 40 of 40 with PXE
                                    Benjamin THIEN Specialist

                                    If ytou want to read the answers of my similar problem, maybe it give you a way...