e00084f8 – The network connection to the Backup Exec Remote Agent has been lost – Check for network errors – refer to link V-79-57344-34040

Problem

Backup job fails with an error:

Final error: 0xe00084f8 – The network connection to the Backup Exec Remote Agent has been lost. Check for network errors.

Error Message

Backup Job Log shows:

Final error: 0xe00084f8 – The network connection to the Backup Exec Remote Agent has been lost. Check for network errors.
Final error category: Resource Errors

For additional information regarding this error refer to link V-79-57344-34040

Solution

[ I ] Check the event viewer, application logs. If getting the below given error, then try creating registry  IRPStackSize as shown below:

 Windows Application Log shows:

Log Name:      Application
Source:        Application Error
Event ID:      1000
Task Category: (100)
Level:         Error
 

Description:
Faulting application name: dllhost.exe, version: 6.1.7600.16385, time stamp: 0x4a5bca54
Faulting module name: xs2.dll, version: 6.0.6001.18000, time stamp: 0x4bf51d4f
Exception code: 0xc0000005
Fault offset: 0x00000000000012c7
Faulting process id: 0x9fc
Faulting application start time: 0x01cca850c561f10f
Faulting application path: C:\Windows\system32\dllhost.exe
Faulting module path: C:\Program Files (x86)\Citrix\XenTools\xs2.dll
Report Id: 0aabd537-1444-11e1-8cbf-36c3f33c5f23

Warning: Symantec technical support strongly recommends that you back up the system registry before making any changes. Incorrect changes to the registry could result in permanent data loss or corrupted files. Please make sure you modify only the keys specified. Please see the document How to back up the Windows registry before proceeding.

Solution:

1. Open Registry on the remote server
2. Start->Run->Regedit
3. Navigate to HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\LanmanServer\Parameter
4. Create a new Dword value named IRPStackSize (case sensitive).
5. Modify the value to between 11 and 50. (15 is the default value when the key is not present; Recommend starting at 30.  Note: On some computers, values from 33 through 38 and above can cause problems.)
6. Reboot the server.

More information about the IRPStackSize parameter can be found on the following Microsoft article:
support.microsoft.com/kb/285089

[ II ] The above mentioned error could happen if limited number of ports are being used to establish connectivity between the media server and the remote server.

Note: It is recommended to keep a range of ports opened instead of just one because a dynamic port can be engaged by any other application and cause can cause data connection issues. 

Therefore keep at least 25 ports opened for the remote system.

The remote agent uses the Network Data Management Protocol (NDMP) and binds to TCP Port 10000 on a remote machine by default. It creates a listener on this port, waiting for the media server to establish a control connection. Once a data move command has been sent across the control connection, the remote agent will initiate a data connection for the Backup Exec job engine. The remote agent and the Backup Exec job engine will utilize TCP ports 1025 to 65000 between the media server and remote.

Solution :

1. Uncheck the option “Enable remote agent TCP dynamic port range” through Tools> Options> Network and Security

Hence by default BE can use ports from 1025 – 65355

OR

1. Specify the ports under the Tools – Options – Network and security- “Enable remote agent TCP dynamic port range”

Open Port 10000 and a port range (minimum of 25 ports) e.g.: 11000 – 11025

In Backup Exec newer releases, Network and Security options can be accessed from Backup Exec Config Icon (top left ) > Configuration and Settings > Backup Exec settings > Network and Security > “Enable TCP dynamic port range”.

[ III ] This error may be received if the Backup Exec Media server and/or Remote Agent servers that are connected to the network are incorrectly configured. In some instances, this could be attributed to piping the Backup Exec Server and/or the Remote Server to a 10Base T Hub and not directly to the switch.

Solution:

1. Run an AppCritical test between the servers to analyze possible packet loss and/or limited connectivity due to incorrect duplex settings, NIC Drivers/Settings, and High Utilization.

See the referenced information below to run an AppCritical test; also, this same information is available from  https://www.veritas.com/support/en_US/article.100023042Download the AppCritical Sequencer from   https://www.appneta.com/sas
 The Sequencer executables must be run with administrative rights.
 Copy the executable into a folder on the server where the network test will be run from.
 Open a command window; at the command line, run:
sas -t <address_of_target_machine>
 If you use gigabit NICs, increase the burst size to 20 in order to get optimal results for high-speed paths:
sas -t <address_of_target_machine> -b 20
 Note: When complete, the test XML file(s) will be created in the “out” directory.
 Email these files back to the support engineer for analysis.
 Note: Whenever possible, target servers, not routers. Also, the target server must be able to respond to ping and traceroute (ICMP) requests.

1. Eliminate the 10Base T hub and connect the media server and/or the remote agent server directly to the switch.
2. Re-run the backup to verify the connection does not drop.

Source: https://www.veritas.com/support/en_US/article.100005195 #kbe00084f8