Backup without name resolution from Veeam Agent to Veeam repository

[Asahi]

HiTeam,

I have a question about name resolution and network routes in Veeam.

First, the Veeam Management Server has 3 IP addresses.
This server is also used as a repository.

At this time, I assume that I will be backing up a Windows physical machine using Veeam Agent.
Protection Group and then create a backup job.

In the backup job, you designate the repository as the Veeam Management Server.
When you run the backup, it tries to make a connection to the Veeam Management & Repository server using the hostname.
In other words, the backup will fail if the Veeam Agent machine => Veeam Management & Repository server is not name resolved.
If the name is resolved, the backup succeeds, but if the name is not resolved, the job fails.

Is this due to the fact that the Veeam server is registered by hostname on the Veeam console of the Veeam Management Server?
※ In the default installation of Veeam, the Veeam server is registered with the hostname "Backup Infrastructure" from the beginning.

For example, if I register the Veeam Management server as an additional Windows machine to the "Backup Infrastructure" with its IP address, and then register the server registered with its IP address as a repository, will the backup succeed even if the Veeam Agent machine is not name resolved to the Veeam Management & Repository server?

In other words, I am looking for a way to back up to the Veeam repository without name resolution from the Veeam Agent machine.

Kind Regards,
Asahi,
Climb Inc.

[HannesK]

Hello,

it tries to make a connection to the Veeam Management & Repository server using the hostname.

that makes no sense to me, if you entered the IP address of the backup server (instead of DNS name). Do you have a case number that backs this statement?

Best regards,
Hannes

[Asahi]

Hello Hannes!!

No, there is no case open.
※I asked this question in this forum because I have not received the log from Customer yet.

Also, the "DNS name or external IP address field" does not appear because the backup job was created with Managed by Backup Server.

Kind Regards,
Asahi,
Climb Inc.

[HannesK]

Hello,
I see the following options

1) fix DNS / network design (a backup server with three network cards is usually a bad design, because that means uncontrolled traffic between networks without a firewall in between). In worst case, hosts file could help
2) Use "managed by agent"
3) you can try out your "hack". It might work, but it's untested from Veeam side.

I would do 1 or 2.

Best regards,
Hannes

[Asahi]

Hi Hannes,

Thank you for reply.

(1) Unfortunately, I don't think it is possible to change the network configuration.

(2) If I use the "managed by agent" method, does this mean that I can back up to the Veeam Management (Repository) server using communication to a specific IP address, without having to resolve names?

Sorry, I have an additional question I would like to know about networking.
When backing up using Veeam Agent for Windows and Veeam Agent for Linux, is there any criteria as to which repository IP address is used for communication to the repository?

When Customer performs a backup with Veeam Agent for Windows, the log file shows some errors in the communication to the repository as follows
===
[18.10.2022 15:53:49] <01> Error Failed to connect to agent's endpoint '192.168.AAA.AAA:2500'. Host: 'hogehoge'.
[18.10.2022 15:53:49] <01> Error The connection could not be established because the connected caller did not respond correctly after a certain time. Or the established connection failed because the connected host did not respond. 192.168.AAA.AAA:2500 (System.Net.Sockets.SocketException)

[18.10.2022 15:54:10] <01> Error Failed to connect to agent's endpoint '192.168.BBB.BBB:2500'. Host: 'hogehoge'.
[18.10.2022 15:54:10] <01> Error The connection could not be established because the connected caller did not respond correctly after a certain time. Or the established connection failed because the connected host did not respond. 192.168.BBB.BBB:2500 (System.Net.Sockets.SocketException)
===

After outputting an error like this, finally the communication to '192.168.CCC.CCC:2500' Host: 'hogehoge' is confirmed and the backup data is transferred.

In the above example, the 'hogehoge' server has three addresses, 192.168.AAA.AAA, 192.168.BBB.BBB, and 192.168.CCC.CCC. Is it possible to perform communication to 192.168.CCC.CCC from the beginning?

Also, for some reason, the backup job of Veeam Agent for Linux is transferring data to 192.168.CCC.CCC from the beginning.

I have not specified any "Preferred Networks" settings on the Veeam Management Server.
Is there any difference in IP address selection between Veeam Agent for Linux and Veeam Agent for Windows?

Kind Regards,
Asahi,
Climb Inc.

[Asahi]

Hi Hannes,

Sorry , I would like to make an additional confirmation about (2).

If I want to use "Managed by Backup Server" and backup from Agent machine to repository without name resolution, is it OK if I register the repository server by IP address and specify it in the backup job?
In this case, I want to register the Veeam backup server as a Windows machine to Backup Infrastructure using its IP address and register it as a repository.

Kind Regards,
Asahi,
Climb Inc.

[HannesK]

Hello,

1) hosts files is not network

After outputting an error like this, finally the communication to '192.168.CCC.CCC:2500' Host: 'hogehoge' is confirmed and the backup data is transferred.

2) looks correct to me. the agent tires everything and once it works, it takes that IP.

Yes, I remember that the Linux agent has a different order. Overall it sounds irrelevant to me, as long as it finally picks a working IP (again, proper network design would solve that error message).

If I want to use "Managed by Backup Server" and backup from Agent machine to repository without name resolution, is it OK if I register the repository server by IP address and specify it in the backup job?

we don't test that setup. But I ask myself, why you want to do it. As you stated in the second post, the backup data is transferred. So everything is working fine.

Best regards,
Hannes

[Ben.online]

In my experience the agents needs to connect to the VBR with name resolution if managed by a Protection Group with preinstalled agents (policy)
The logs says it is failing over to ip and the ip is accesable pingable and telnet to ip:10005 works fine.

[27.10.2022 15:42:04] <01> Error Failed to resolve connection point XXXXX name to ip addresses.
[27.10.2022 15:42:04] <01> Error Failed to resolve host name 'XXXXX' (Veeam.Backup.Common.CDnsException)
[27.10.2022 15:42:04] <01> Error at Veeam.Backup.Common.DnsEx.GetIPAddresses(String hostNameOrAddress)
[27.10.2022 15:42:04] <01> Error at Veeam.Backup.Common.CIpAddress.TryResolveConnectionPoint(String ipOrDnsName, IPAddress[]& ipAddresses)
[27.10.2022 15:42:04] <01> Info Failover to ips: 192.168.XXX.X 192.168.XX.XX 192.168.XXX.X 192.168.XXX.XX 192.168.XXX.X 192.168.XXX.XX
[27.10.2022 15:42:04] <01> Info Personal agent certificate with thumbprint XXXX9418CF03DC5093B4D92949F59F0B7AFB6 found.
[27.10.2022 15:43:04] <01> Error Failed to perform the operation. (Veeam.EndPoint.Configurator.ExecutionError)
[27.10.2022 15:43:04] <01> Error at Veeam.EndPoint.Configurator.Program.Perform(ExecutionManager manager)
[27.10.2022 15:43:04] <01> Error at Veeam.EndPoint.Configurator.Program.ExecuteInner(String[] rawCommandLine, CInputXmlData outputData, ServiceHandler serviceHandler, CInputXmlData executionResult, INotificationOutput& notificationOutput, ExecutionArgs& args, List`1& messages, List`1& warnings, List`1& amMessages, ExecutionOutput& dataOutput)
[27.10.2022 15:43:04] <01> Error at Veeam.EndPoint.Configurator.Program.Execute(String[] rawCommandLine, Mutex executionMutex)
[27.10.2022 15:43:04] <01> Error Exception has been thrown by the target of an invocation. (System.Reflection.TargetInvocationException)

But the agents are not connecting to the VBR until i add the VBR in the hostfile of the agent machine
Then once the initial setup/sync is completed the agent will use the ip for backup.
After removing the name from hosts file backups will work fine
But policy changes and resync/scan will stop, you can find the following entries in the Config.ExportImport.log :

[02.11.2022 12:21:24] <01> Info [ExecutionManager] Initialization.
[02.11.2022 12:21:24] <01> Info [ExecutionManager] Choice performer for execution mode: StartSelfSync.
[02.11.2022 12:22:26] <01> Error Failed to perform the operation. (Veeam.EndPoint.Configurator.ExecutionError)
[02.11.2022 12:22:26] <01> Error at Veeam.EndPoint.Configurator.Program.Perform(ExecutionManager manager)
[02.11.2022 12:22:26] <01> Error at Veeam.EndPoint.Configurator.Program.ExecuteInner(String[] rawCommandLine, CInputXmlData outputData, ServiceHandler serviceHandler, CInputXmlData executionResult, INotificationOutput& notificationOutput, ExecutionArgs& args, List`1& messages, List`1& warnings, List`1& amMessages, ExecutionOutput& dataOutput)
[02.11.2022 12:22:26] <01> Error at Veeam.EndPoint.Configurator.Program.Execute(String[] rawCommandLine, Mutex executionMutex)
[02.11.2022 12:22:26] <01> Error Exception has been thrown by the target of an invocation. (System.Reflection.TargetInvocationException)
[02.11.2022 12:22:26] <01> Error at System.RuntimeMethodHandle.InvokeMethod(Object target, Object[] arguments, Signature sig, Boolean constructor)
[02.11.2022 12:22:26] <01> Error at System.Reflection.RuntimeMethodInfo.UnsafeInvokeInternal(Object obj, Object[] parameters, Object[] arguments)
[02.11.2022 12:22:26] <01> Error at System.Delegate.DynamicInvokeImpl(Object[] args)
[02.11.2022 12:22:26] <01> Error at Veeam.Common.Remoting.CRetryableMethod.DoInvoke(Delegate dlg, Object[] args)
[02.11.2022 12:22:26] <01> Error at Veeam.Backup.Interaction.Management.Client.CJobManagementService.PerformSelfDiscover()
[02.11.2022 12:22:26] <01> Error at Veeam.Agent.Configurator.CStartSelfSyncPerformer.Perform()
[02.11.2022 12:22:26] <01> Error at Veeam.EndPoint.Configurator.Program.Perform(ExecutionManager manager)
[02.11.2022 12:22:26] <01> Error Exception has been thrown by the target of an invocation. (System.Reflection.TargetInvocationException)
[02.11.2022 12:22:26] <01> Error Server stack trace:
[02.11.2022 12:22:26] <01> Error at Veeam.Backup.ForeignBackupServiceClient.CForeignBackupServiceClientFactory.Connect(CForeignServerConnectionInfo foreignConnInfo, String sessionName, EForeignServiceSessionType sessionType, String localMachineName, Boolean persistentConnection, Boolean keepLeaseAlive, Boolean forceUseSsl)
[02.11.2022 12:22:26] <01> Error at Veeam.Backup.Core.CForeignBackupService.Connect(CForeignServerConnectionInfo connParams, String sessionName, EForeignServiceSessionType sessionType)
[02.11.2022 12:22:26] <01> Error at Veeam.EndPoint.Core.CCatchAllSelfDiscoverManager.PerformInternal(Boolean isPoliciesApplyRequired)
[02.11.2022 12:22:26] <01> Error at Veeam.EndPoint.Core.CConfigurationUpdater.PerformSelfDiscover()
[02.11.2022 12:22:26] <01> Error at System.Runtime.Remoting.Messaging.StackBuilderSink._PrivateProcessMessage(IntPtr md, Object[] args, Object server, Object[]& outArgs)
[02.11.2022 12:22:26] <01> Error at System.Runtime.Remoting.Messaging.StackBuilderSink.SyncProcessMessage(IMessage msg)
[02.11.2022 12:22:26] <01> Error Exception rethrown at [0]:
[02.11.2022 12:22:26] <01> Error at System.Runtime.Remoting.Proxies.RealProxy.HandleReturnMessage(IMessage reqMsg, IMessage retMsg)
[02.11.2022 12:22:26] <01> Error at System.Runtime.Remoting.Proxies.RealProxy.PrivateInvoke(MessageData& msgData, Int32 type)
[02.11.2022 12:22:26] <01> Error at Veeam.Backup.Interaction.Management.IJobManagementService.PerformSelfDiscover()
[02.11.2022 12:22:26] <01> Error Failed to connect to remote backup service. (System.TimeoutException)

If u use unmanaged agents or agents as individual computers in a protection group than name resolution is not used

I know the network setup is not ideal nor is editing the hosts file, but having a shared DNS containing the backup infrastructure over multiple customers is even less