rudder-agent 3.1 : different output between nodes

Question

I'm using rudder-agent 3.1 (rudder server 3.0.6), when I'm forcing inventory with "rudder agent inventory" I don't have the same output size between node, it's like the verbose was different (see bottom).

Any idea ?

x@pv2-51b7y:~$ sudo rudder agent inventory
2015-07-29T09:47:23+0200   notice: /default/doInventory/methods/'any'/default/fusionAgent/commands/'/opt/rudder/bin/run-inventory --local=/var/rudder/tmp/inventory/pv2-51b7y-e5f0d268-0d81-4f27-98de-edee40604b0e.ocs --scan-homedirs'[0]: Q: "...-inventory --lo": [info] FusionInventory instance: builtin
Q: "...-inventory --lo": [info] Inventory saved in /var/rudder/tmp/inventory/pv2-51b7y-e5f0d268-0d81-4f27-98de-edee40604b0e.ocs
Q: "...-inventory --lo": [info] task WakeOnLan execution not requested
Q: "...-inventory --lo": [info] task NetDiscovery execution not requested
Q: "...-inventory --lo": [info] task NetInventory execution not requested

x@pv2-51b7y:~$

and

x@pv2-8kmcm:~$ sudo rudder agent inventory
R: @@Common@@log_info@@hasPolicyServer-root@@common-root@@00@@common@@StartRun@@2015-07-29 07:46:27+00:00##8cc35d59-bd4b-4fab-bd37-4ec837743220@#Start execution
R: @@Common@@result_success@@hasPolicyServer-root@@common-root@@00@@Security parameters@@None@@2015-07-29 07:46:27+00:00##8cc35d59-bd4b-4fab-bd37-4ec837743220@#The internal environment security is acceptable
R: @@Common@@result_success@@hasPolicyServer-root@@common-root@@00@@Process checking@@None@@2015-07-29 07:46:27+00:00##8cc35d59-bd4b-4fab-bd37-4ec837743220@#There is an acceptable number of cf-execd processes (between 0 and 2) and cf-agent processes (between 0 and 5)
R: @@Common@@result_success@@hasPolicyServer-root@@common-root@@00@@CRON Daemon@@None@@2015-07-29 07:46:27+00:00##8cc35d59-bd4b-4fab-bd37-4ec837743220@#The CRON daemon is running
R: @@Common@@result_success@@hasPolicyServer-root@@common-root@@00@@Binaries update@@None@@2015-07-29 07:46:27+00:00##8cc35d59-bd4b-4fab-bd37-4ec837743220@#The CFengine binaries in /var/rudder/cfengine-community/bin are up to date
R: @@Inventory@@log_info@@inventory-all@@inventory-all@@00@@inventory@@None@@2015-07-29 07:46:27+00:00##8cc35d59-bd4b-4fab-bd37-4ec837743220@#An inventory was already sent less than 8 hours ago
2015-07-29T09:46:36+0200   notice: /default/doInventory/methods/'any'/default/fusionAgent/commands/'/opt/rudder/bin/run-inventory --local=/var/rudder/tmp/inventory/pv2-8kmcm-8cc35d59-bd4b-4fab-bd37-4ec837743220.ocs --scan-homedirs'[0]: Q: "...-inventory --lo": [info] FusionInventory instance: builtin
Q: "...-inventory --lo": [info] Inventory saved in /var/rudder/tmp/inventory/pv2-8kmcm-8cc35d59-bd4b-4fab-bd37-4ec837743220.ocs
Q: "...-inventory --lo": [info] task WakeOnLan execution not requested
Q: "...-inventory --lo": [info] task NetDiscovery execution not requested
Q: "...-inventory --lo": [info] task NetInventory execution not requested

R: 

@@Inventory@@log_debug@@inventory-all@@inventory-all@@00@@inventory@@None@@2015-07-29 07:46:27+00:00##8cc35d59-bd4b-4fab-bd37-4ec837743220@#Running inventory
R: @@Inventory@@log_info@@inventory-all@@inventory-all@@00@@inventory@@None@@2015-07-29 07:46:27+00:00##8cc35d59-bd4b-4fab-bd37-4ec837743220@#Curl installed
R: @@Inventory@@log_debug@@inventory-all@@inventory-all@@00@@inventory@@None@@2015-07-29 07:46:27+00:00##8cc35d59-bd4b-4fab-bd37-4ec837743220@#Generated the CPUID
R: @@Inventory@@log_debug@@inventory-all@@inventory-all@@00@@inventory@@None@@2015-07-29 07:46:27+00:00##8cc35d59-bd4b-4fab-bd37-4ec837743220@#The user list generation tool is not present yet. Skipping...
R: @@Inventory@@result_success@@inventory-all@@inventory-all@@00@@inventory@@None@@2015-07-29 07:46:27+00:00##8cc35d59-bd4b-4fab-bd37-4ec837743220@#The inventory has been successfully sent
R: @@Common@@log_info@@hasPolicyServer-root@@common-root@@00@@common@@EndRun@@2015-07-29 07:46:27+00:00##8cc35d59-bd4b-4fab-bd37-4ec837743220@#End execution
x@pv2-8kmcm:~$

Answer 1

The first output doesn't exhibit any reports, while the second is (StartRun, EndRun, and in between you have success reports)

You have most probably configured your nodes to use the "Change Only" reports mode.

This mode is designed to minimize number of reports sent over the network, by not sending success nor log reports, and sending an "heartbeat" (start and end run report) once in a while

Your first node is clearly in this mode (no start report, nor end report, but an inventory is already done), while the second is a fresh install (the number between @@ @@ is 0, meaning it's is applying the rules we ship with the agent), and so outputs and sends all the reports it generates.

When you'll accept the second node within Rudder, it will get new rules, and apply the change only report mode