Client computers and Task Servers fail to register and stop receiving tasks after a Data Connector “Edit function” runs on ITMS 8.8

Products

• IT Management Suite
• Client Management Suite
• Server Management Suite
• Asset Management Suite

Issue/Introduction

Client computers take a long time to receive tasks or do not receive them at all. Task Servers stop accepting client registrations. Restarting the Symantec Management Agent on clients or restarting the Task Server services on site servers does not help. Restarting the Notification Server services and IIS on the SMP Server restores task delivery for a few days, then the issue returns.

The Task Server agent logs show entries like:

Operation 'Direct: Head' failed. 
Url: HTTPS://smpserver.example.com:443/Altiris/TaskManagement/ClientTask/ExecTSMethod.aspx?clientTaskServerGuid=d5############43a 
Connection path: 1* - Direct: [10.##.##.22] -> smpserver.example.com [10.##.##.22:443] 
Communication profile id: {45###############5c} 
Throttling: 0 0 0 
Connecton stage: DNS resolve 
Error type: HTTP error 
Error code: HTTP status 503: The service is temporarily overloaded (0x8FA101F7) 
Error note: 503 Service Unavailable 
Server SSL connection info: 
   Server certificate: 
      Serial number: 7d ############# d7 
      Thumbprint: a8 ##############a 8f 
   Client certificate: 
      Serial number: N/A 
      Thumbprint: N/A 
   Cryptographic protocol: TLS 1.2 
   Cipher suite: TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384 
   Cipher algorithm: AES 
   Cipher key length: 256 
   Hash algorithm:  
   Hash length: 0 
   Key exchange algorithm: ECDH 
   Key length: 384 
Client SSL attributes for server connection: 
   Client certificate: 
      Serial number: N/A 
      Thumbprint: N/A 
   Cryptographic protocol: TLS 1.1, 1.2
Connection path id: 8b###############fc
-----------------------------------------------------------------------------------------------------
Date: 9/24/2026 11:03:38 AM, Tick Count: 663305078 (7.16:15:05.0780000), Size: 1.51 KB
Process: AtrsHost.exe (8336), Thread ID: 4084, Module: AeXNetComms.dll
Priority: 1, Source: NetworkOperation

Resulting symptoms on the SMP Server:

1. AtrsHost.exe and w3wp.exe fail to open interprocess communication streams to AeXSvc.
2. License checks in GetClientPolicies calls from w3wp.exe time out.
3. Clients report that no new policies are available.
4. The Client Task Agent cannot register with any Task Server.

Example NS log warning entry:

Initialization failed for Producer stream init to [aexsvc], source=[atrshost]

Consumer process has terminated.
   [EndOfStreamException @ Altiris.GroupCommunications.dll]
   at Altiris.GroupCommunications.Support.GCInterprocessStream.CreateViewAndService(InterprocessStreamMode)
   at Altiris.GroupCommunications.Support.GCInterprocessStream.CreateMmfAndService()
   at Altiris.GroupCommunications.Support.GCInterprocessStream.TryInitialize(GCMutex, out int)
   at Altiris.GroupCommunications.Support.GCInterprocessStream.Initialize(Address, InterprocessStreamMode, int, WaitHandle[], ReliableMmfProtocol)

Exception logged from:
   at Altiris.Profiling.Support.LogSupport.BuildExceptionMessage(string, Exception, string, bool)
   at Altiris.GroupCommunications.Support.GCTrace.TraceRegardless(TraceLevel, Exception, string)
   at Altiris.GroupCommunications.Support.GCInterprocessStream.Initialize(Address, InterprocessStreamMode, int, WaitHandle[], ReliableMmfProtocol)
   at Altiris.GroupCommunications.Support.GCInterprocessStream.Produce(Address, int, WaitHandle[], ReliableMmfProtocol)
   at Altiris.GroupCommunications.Protocols.ReliableMmfProtocol+MmfTransport.InitializeStream(ref EMmfWriteFailReason)
   at Altiris.GroupCommunications.Protocols.ReliableMmfProtocol+MmfTransport.WriteData(Address, byte[], ref EMmfWriteFailReason)
   at Altiris.GroupCommunications.Protocols.ReliableMmfProtocol+MmfTransport.WriteData(Address, byte[], int, ref EMmfWriteFailReason)
   at Altiris.GroupCommunications.Protocols.ReliableMmfProtocol+MmfTransport.ProcessQueueWorkItemCallback(object)
   at Altiris.Common.Threading.BalancedThreadPool.ExecuteWorkerRequest(BalancedThreadPoolWorkerState, BalancedThreadPoolWorkerRequest)
   at Altiris.Common.Threading.BalancedThreadPool.ThreadPoolProc(object)
   at System.Threading.ThreadHelper.ThreadStart(object)

-----------------------------------------------------------------------------------------------------
Date: 9/24/2026 10:04:30 AM, Tick Count: 1750943265 (20.06:22:23.2650000), Size: 2.13 KB
Process: AtrsHost (8820), Thread ID: 11, Module: Altiris.NS.dll
Priority: 2, Source: GCInterprocessStream.Initialize

 

Initialization failed for Producer stream init to [aexsvc], source=_lm_w3svc_1_root_altiris_taskmanagement_2_134346517498235537_2d97758054774fadaedad29f6c4a983a

Consumer process has terminated.
   [EndOfStreamException @ Altiris.GroupCommunications.dll]
   at Altiris.GroupCommunications.Support.GCInterprocessStream.CreateViewAndService(InterprocessStreamMode)
   at Altiris.GroupCommunications.Support.GCInterprocessStream.CreateMmfAndService()
   at Altiris.GroupCommunications.Support.GCInterprocessStream.TryInitialize(GCMutex, out int)
   at Altiris.GroupCommunications.Support.GCInterprocessStream.Initialize(Address, InterprocessStreamMode, int, WaitHandle[], ReliableMmfProtocol)

Exception logged from:
   at Altiris.Profiling.Support.LogSupport.BuildExceptionMessage(string, Exception, string, bool)
   at Altiris.GroupCommunications.Support.GCTrace.TraceRegardless(TraceLevel, Exception, string)
   at Altiris.GroupCommunications.Support.GCInterprocessStream.Initialize(Address, InterprocessStreamMode, int, WaitHandle[], ReliableMmfProtocol)
   at Altiris.GroupCommunications.Support.GCInterprocessStream.Produce(Address, int, WaitHandle[], ReliableMmfProtocol)
   at Altiris.GroupCommunications.Protocols.ReliableMmfProtocol+MmfTransport.InitializeStream(ref EMmfWriteFailReason)
   at Altiris.GroupCommunications.Protocols.ReliableMmfProtocol+MmfTransport.WriteData(Address, byte[], ref EMmfWriteFailReason)
   at Altiris.GroupCommunications.Protocols.ReliableMmfProtocol+MmfTransport.WriteData(Address, byte[], int, ref EMmfWriteFailReason)
   at Altiris.GroupCommunications.Protocols.ReliableMmfProtocol+MmfTransport.ProcessQueueWorkItemCallback(object)
   at Altiris.Common.Threading.BalancedThreadPool.ExecuteWorkerRequest(BalancedThreadPoolWorkerState, BalancedThreadPoolWorkerRequest)
   at Altiris.Common.Threading.BalancedThreadPool.ThreadPoolProc(object)
   at System.Threading.ThreadHelper.ThreadStart(object)

-----------------------------------------------------------------------------------------------------
Date: 9/24/2026 10:48:41 AM, Tick Count: 1753594281 (20.07:06:34.2810000), Size: 2.25 KB
Process: w3wp (8848), Thread ID: 731, Module: Altiris.NS.dll
Priority: 2, Source: GCInterprocessStream.Initialize

Environment

Product:	IT Management Suite (ITMS) 8.8
Fixed in:	ITMS 8.8.1 and later
Point fix:	Point fix for 8.8 release
Components:	Symantec Management Platform (SMP), Data Connector (data sources with “Pre-process import data” enabled), Altiris Service (AeXSvc.exe), Altiris Object Host Service (AtrsHost.exe), IIS (w3wp.exe), Task Server, Client Task Agent, Symantec Management Agent

Cause

A Data Connector data source can run custom C# code through the “Pre-process import data” option (“Edit function”).


On ITMS 8.8, Data Connector runs this code in an extra .NET application domain inside the Altiris Service process (AeXSvc.exe). When that application domain unloads, it fires an application-wide shutdown event. That event disposes several AeXSvc components, including Group Communications, while the service itself keeps running.

After this happens, the Object Host Service (AtrsHost.exe) and IIS worker processes (w3wp.exe) can no longer communicate with AeXSvc. License checks during client policy requests time out. Clients stop receiving new policies, and the Client Task Agent cannot register with any Task Server, local or remote.

The custom code itself is not the cause. The way Data Connector executes and unloads it is the cause. The fix is a change in the Altiris.Profiling module, included in ITMS 8.8.1.

Resolution

The defect is fixed in ITMS 8.8.1 release. Broadcom Dev team produced and validated a point fix for those in ITMS 8.8 release.

In order to obtain the available pointfix for ITMS 8.8 release, refer to: CUMULATIVE POST ITMS 8.8 RTM(GA) POINT FIXES

The following steps are provided as a guidance on how to validate the issue and what can be done in the meantime:

Part 1: Recover service - Current Workaround to get things working for the moment

1. On the SMP Server, stop the Altiris services, then stop the World Wide Web Publishing Service.
2. Start the World Wide Web Publishing Service and the Altiris services again. The services restarted in this case were:

• • Altiris Service
  • Altiris Client Message Dispatcher
  • Altiris Client Task Data Loader
  • Altiris Object Host Service
  • IIS

This restores task delivery temporarily. The issue returns the next time a data source with an Edit function runs.

Part 2: Find data sources that use an Edit function

1. In the Symantec Management Console, navigate to:
   Settings > All Settings > Notification Server > Connector > Data Sources
2. Open each data source in the list.
3. Check whether the "Pre-process import data" check box is selected.
4. Click "Edit function..." to view the code snippet. Snippets use importData as the default object name for the imported data.
5. Record each data source that has Pre-process import data selected, and the import rules that use it.

Part 3: Apply the workaround

1. On each data source found in Part 2, clear the Pre-process import data check box.
2. Click Save changes.
3. Filter the import data by another method until the permanent fix is in place.
4. Restart the services as described in Part 1.

Do not use processing snippets in Data Connector until the SMP Server is on 8.8.1 or later, or has the 8.8 point fix.

Part 4: Apply the permanent fix

1. Upgrade to ITMS 8.8.1 or later.
2. If an upgrade is not possible, get the pointfix mentioned at the beginning of this KB article. 

Part 5: If the issue returns with no Edit function in use

Collect evidence while the problem is active, before restarting any service. Once the services restart, this evidence is gone. 

Before the next occurrence:

1. Confirm that the antivirus exclusions in KB Recommended Anti-virus exclusions for Symantec Management Platform and Agents are applied on the SMP Server. Ask the security vendor whether behavior, memory, or exploit protection features apply to AeXSvc.exe, AtrsHost.exe, and w3wp.exe. File-scan exclusions do not always cover these features.
2. Increase the SMP log file size and file count (KB Configuring Advanced Logging for Notification Server and Symantec Management Agent). During this problem the log rolls over about every 30 seconds and keeps about 13 minutes of history.
3. Download Microsoft Sysinternals ProcDump and Handle to C:\temp\ITMSsupport\Tools.
4. Make sure the target drive has free space for full memory dumps of AeXSvc.exe, AtrsHost.exe, and all w3wp.exe processes.

When the problem returns, run these commands from one elevated PowerShell window:

1. Set the folder variables:

--powershell

$dir   = 'C:\temp\ITMSsupport'
$tools = "$dir\Tools"
$dumps = "$dir\Dumps"
New-Item -Path $dir,$tools,$dumps -ItemType Directory -Force | Out-Null

2. Write down the time the problem was noticed and anything that ran shortly before it, such as security scans, definition updates, Windows updates, or backups. Save the notes as C:\temp\ITMSsupport\notes.txt.
   
   
3. Copy the SMP logs:

--powershell

Copy-Item 'C:\ProgramData\Symantec\SMP\Logs' -Destination "$dir\SMP_Logs" -Recurse

4. Record process IDs, start times, and IIS application pools:

--powershell

Get-Process AeXSvc,AtrsHost,w3wp | Select-Object Name,Id,StartTime | Out-File "$dir\process_list.txt"
& "$env:windir\system32\inetsrv\appcmd.exe" list wp | Out-File "$dir\w3wp_app_pools.txt"

5. Capture full memory dumps:

--powershell

& "$tools\procdump64.exe" -accepteula -ma AeXSvc.exe $dumps
& "$tools\procdump64.exe" -accepteula -ma AtrsHost.exe $dumps
Get-Process w3wp | ForEach-Object { & "$tools\procdump64.exe" -accepteula -ma $_.Id $dumps }

6. List the interprocess communication handles:

--powershell

& "$tools\handle64.exe" -accepteula -a -p AeXSvc.exe ips | Out-File "$dir\handles_aexsvc.txt"
& "$tools\handle64.exe" -a -p AtrsHost.exe ips | Out-File "$dir\handles_atrshost.txt"
& "$tools\handle64.exe" -a -p w3wp.exe ips | Out-File "$dir\handles_w3wp.txt"
& "$tools\handle64.exe" -a ipsWriterMutex | Out-File "$dir\handles_all_ipsWriterMutex.txt"

AeXSvc should own handles named Global\ipsWriterMutex<GUID>, Global\ipsWriterSemaphore<GUID>, and Global\ipsReaderSemaphore<GUID>. Check whether they are missing or held by another process. If a file is empty, run the same command without ips at the end.

7. List loaded modules:

--powershell

tasklist /m /fi "IMAGENAME eq AeXSvc.exe"   | Out-File "$dir\modules_aexsvc.txt"
tasklist /m /fi "IMAGENAME eq AtrsHost.exe" | Out-File "$dir\modules_atrshost.txt"
tasklist /m /fi "IMAGENAME eq w3wp.exe"     | Out-File "$dir\modules_w3wp.txt"

8. Export the Windows event logs:

--powershell

wevtutil epl System "$dir\System.evtx"
wevtutil epl Application "$dir\Application.evtx"

Also export security product event or block logs for the 24 hours before the problem.

9. Confirm that every file has a non-zero size and that Dumps has one .dmp file per process:

--powershell

Get-ChildItem $dir -Recurse -File | Select-Object FullName,Length

10. Restart the services as described in Part 1.
11. Upload the C:\temp\ITMSsupport folder to the support case.

Diagnostics

Diagnostic step	Command / path	Expected result	Actual result	Interpretation
Check SMP Server logs for interprocess stream failures	C:\ProgramData\Symantec\SMP\Logs (Log Viewer, source GCInterprocessStream.Initialize)	No “Initialization failed for Producer stream” entries	Repeated entries from AtrsHost and w3wp; log flooded	AtrsHost.exe and w3wp.exe cannot reach AeXSvc through Group Communications
Check SMP Server logs for license check failures	C:\ProgramData\Symantec\SMP\Logs	No ThreadAbortException in LicensingService.CheckLicense	Entries from w3wp.exe	GetClientPolicies calls time out while checking licenses through AeXSvc
Check Task Server agent log	Symantec Management Agent logs on the site server (C:\ProgramData\Symantec\Symantec Agent\Logs)	Task Server registers with the SMP Server	HTTP 503 on ExecTSMethod.aspx; Task Server in paused state	Task Server cannot complete registration because the SMP Server does not respond
Check client agent log	Symantec Management Agent logs on the client (Symantec Management Agent logs on the site server (C:\ProgramData\Symantec\Symantec Agent\Logs)	Client Task Agent registers with a Task Server	Element not found (0x80070490) on registration	Client cannot get a Task Server; the fault is on the SMP Server, not the Task Server
Check Data Connector data sources	Settings > All Settings > Notification Server > Connector > Data Sources	Pre-process import data cleared, or SMP Server on 8.8.1+ release	Edit function in use on 8.8	Matches the confirmed trigger
Check NS and SQL Server time	Compare system time on both servers	Times match	About 20 seconds offset in this case	Dev noted this is not the cause but is worth correcting