Symantec Management Platform (Notification Server)

Hello,

i have like 10% of the computers for wich i don't have inventory in the Notification Server, everything seems ok at the client level.
What i did:
- uninstall/install inventory plugin
- uninstall/install Symantec agent
- resetguid of the agent

i tried what is explain in this link: http://www.symantec.com/business/support/index?page=content&id=HOWTO32482

force the client to write the inventory to a file, and dropped that file in EvtQFast folder, the file is generated and i can see the hardware inventory in the file, and the file dissapear from EvtQFast folder so i suppose its processed since i have no error in the log viewer.

But still no info in NS

Any ideas about how i could fix this?

regards

You unchecked "send deltas", right?  When you say you don't have inventory for these 10%, what are you seeing?  I assume you mean that 100% of Operating System, Hardware, and Software data classes are 100% empty.  Is that what you're saying?

You may wish to check licensing in SIM.  Do you have sufficient licenses?

What do you see in the Altiris Log Viewer, if anything, when the full inventory is processed from the client?  I usually test with a task that runs a full (not delta!) hardware inventory, since this takes about 5 minutes.  What do you see in the Altiris Log Viewer 0-15 seconds after the client reports (via Task History) a success?

You may especially want to look for 'failed to process' or 'failed to insert' messages, as this can represent a problem entering inventory.  Alternatively, check your NS's event queues for NSEs in the Bad folders, which could indicate malformed NSE's.

Hello,

thanks for your reply

Yes i have unchecked "send deltas"

actually i have no hardware inventory for +/- 150 computers of a total of 1700.

the client generate the inventory file since i changed the registry parameter so it writes it to a file and when opening it i can see the hardware information. The task is in success in the task history on the SMP and on the client side.

When i copy the nse file to a evtQFast folder, it dissapears directly from the folder, and nothing stays in the bad folder, i I don't see anything special in the log viewer, only 2 messages saying it failed to move the file but i assume it is during the copy of the file.

<event date="Feb 29 14:58:37 +00:00" severity="2" hostName="SP0097" source="MoveFile" module="AeXSVC.exe" process="AeXSvc" pid="1496" thread="313" tickCount="769118375"><![CDATA[Failed to move file from 'E:\Program Files\Altiris\Notification Server\NSCap\EvtQFast\NSE-F9B-00000003-00000047.tmp' to 'E:\Program Files\Altiris\Notification Server\NSCap\EvtQFast\Process\NSE-F9B-00000003-00000047.tmp'. Exception details: System.IO.IOException: The process cannot access the file because it is being used by another process.
   at System.IO.__Error.WinIOError(Int32 errorCode, String maybeFullPath)
   at System.IO.File.Move(String sourceFileName, String destFileName)
   at Altiris.NS.EventRouter.MoveFile(String source, String destination, Int32 severity)]]></event>
<event date="Feb 29 14:58:37 +00:00" severity="2" hostName="SP0097" source="MoveFile" 

so i uninstall the client, remove the computer ressource via the NS Portal, reinstall the agent, than it works, but it will be difficult to do that for 100 clients unless there's a possibility to remove ressources via script/sql

Does that error occur during an inventory that is copied up to the NS by the agent, as opposed to manually by you?

You might want to right-click a data class (e.g. Chassis) and choose Resource Manager or Properties, then record the GUID for the data class.  This is the InventoryClassGuid.

Then right-click your test resource, choose Resource Manager, and record the GUID.  This is the ResourceGuid.

Then run this query:

DELETE FROM ResourceUpdateSummary
WHERE ResourceGuid='ResourceGuid you recorded above'
AND InventoryClassGuid='GUID you recorded above'

Then run your inventory task and allow the agent to copy it up to the server. Is it successfully processed for the Chassis data class?

Hello,

Does that error occur during an inventory that is copied up to the NS by the agent, as opposed to manually by you?

Right

It seems you pointed me to the right direction with the ResourceUpdateSummary table. I made two tests tonight, first i deleted all rows for one clients then rerun a gather inventory tasks, and voila the inventory appears :) but the symantec agent acts a bit weird when rebooting the computer, it was not showing in systray, so i looked for a more smooth solution, on a second buggy agent i only delete the row for the INV_ComputerSystem and rerun the inventory task, and it appears correctly.

So i have a method to correct the computers showing an unknown computer model but other inventory will possibly won't be correct, so i could delete all rows for inventory items but is it a good solution?

Because i on't want to have to check if all inventory tasks are going ok every three months....

mclemson: what's your advice

Only delete for the correct InventoryClassGuids, that is, the ones with problems.  If you just delete every record for a client from the ResourceUpdateSummary table, you will have problems.

I would delete the DataHash values for all InventoryClassGuids that are represented by a full inventory, if you have concerns.  This will force a heavier-than-usual load next time inventory is processed.  This will permanently resolve the issue unless you have SQL deadlocks or other causes that will cause the issue to recur.

So i deleted all ClassGuid corresponding to those tableName, for the computers returning an unkown model in the inventory:

Inv_HW_Baseboard
Inv_HW_Battery
Inv_HW_Chassis
Inv_HW_Computer_System
Inv_HW_Desktop_Monitor
Inv_HW_Disk_Partition
Inv_HW_Display_Controller
Inv_HW_Keyboard
Inv_HW_Logical_Device
Inv_HW_Logical_Disk
Inv_HW_Logical_disk_based_on_partition
Inv_HW_Network_Adapter
Inv_HW_Physical_Memory
Inv_HW_Physical_Memory_Array
Inv_HW_Physical_Memory_Location
Inv_HW_Plug_and_Play_Entity_Windows
Inv_HW_Pointing_Device
Inv_HW_Printer
Inv_HW_Printer_Windows
Inv_HW_Processor
Inv_HW_Processor_Name_Windows
Inv_HW_SCSI_Controller
Inv_HW_Sound_Device
Inv_HW_Storage
Inv_HW_Storage_To_Disk_Partition
Inv_HW_USB_Device
Inv_HW_Video_Controller_Windows

and rerun a gather inventory, so now they appears correctly. Strange thing is that in the ResourceUpdateSummary table the ModifiedDate column was filled with recents date so it seems to run but was doing nothing....

I have some troubles to keep the whole altiris platform running well update after update, i find clients with outdated plugin, inventory not working, computers moved to others ou sometimes synchrnize strangly....

Do you have some advices for smp maintenance that should be done so the platform will run smoothly?

Thanks again for your help

I am seeing the same issue, of about 550 computers I’m not getting inventory for about 30+ computers.  I initially had to create a quick task to push the inventory plug-in and run gather hardware only and still the computers came back as unknown.  I can see that inventory is being gathered by opening up the files under Inventory\NSI but it’s not reflecting in the console or reports.  In the console I see no hardware information appearing and using IT analytic cubes they return as unknown. My solution was to completely remove the Symantec Management Agent and this has solved some of them, but not all.

I agree after these Service Packs I feel like I have to check everything twice.  I also have problems with outdated plug-ins, also the filters for the plug-ins are not matching with each other…  

This is the sql i used to resolve the hardware inventory issue i first get the list of unkown computer from It Analytics than run the query against the NS Database

delete
FROM ResourceUpdateSummary
WHERE [InventoryClassGuid] in
    (select distinct b.[Guid]
        from dbo.DataClass b
        where b.[DataTableName] like 'Inv_HW%')
and ResourceGuid in
    (SELECT b.[Guid]
        FROM [vComputer] b
        where lower(Name) in
('computer1','computer2',etc...))

for outdated plugin i use custom inventory to get fileversion of each symantec agent components and i track old ones who doesn't update normally, my feelings is that if i don't double/triple check the agents then the control of agents will degrade slowly...

Btw, i open the forum discussion and a case on symatec support at the same time, and i still have no reply for my case...

Regards

This is not a service pack issue.  If you have restored the DB from backups, or if you have had problems with SQL deadlocks, the DataHash can be out of date.  The procedure I've shared will allow you to resolve the issue.

Outdated plug-in: Ensure your schedule includes more than just Run Once ASAP.  Add a scheduled time for 00:00 in the past (e.g. yesterday) and to check during the window as well (no more frequent than 1.5x your client policy update interval).  I do every 6 hours usually.

Give this one day to shake out, then resolve every single one that remains.  Do this monthly.  RAADv2 is your friend for tracking down these systems.  Load them up by selecting the filter, then analyze each looking for patterns like a failed task agent, client not getting policies, etc.  Resolve remotely and using automated methods when many agents are involved.

Inventory not working: Continue to work with Symantec to determine the cause of your outdated DataHash values.  SQL deadlocks are one cause.  Restoring from a backup is another.  But this is not specific to this Service Pack release.  Resolve 100% of current cases thoroughly and then monitor to ensure the issue does not reoccur.  If it reoccurs, you know you have an environmental issue to resolve.  By resolving all of them, you know you have a good baseline to start from.

Computer OU synchronization: Tough to tell.

Glad I could help you with the inventory issue.  Please be sure to mark my post as the solution.