VMWARE SAP REPLICATION JOB ERROR

[AdrianDiaz]

Hello everyone, we are starting in my office with a DRP for SAP HANA, we have several problems because the incremental backups can take up to 15 hours to perform and replication jobs more than 25 hours and often show errors like those attached to the topic.

the virtual machine has a size of 1 TB and a 1.5 TB ISCSI for instance and DB backups. can you recommend me some best practices for this?

Regards!

[Andreas Neufert]

Hi Adrian,
are you saying that you have issues with the current Veeam backups and replication jobs?

Anyway can you please share more details?
What Replication type used?
What would you like to achieve overall without thinking about technical limitations? What is the business need?
Is it InGuest iSCSI mounted into the OS?

[PetrM]

Hello,

Further to the previous post, I think it's worth checking html-reports or job session statistics, in particular to look for an amount of data read ("Read") from the production datastore. For instance, the size and duration of incremental backup would be comparable to the full one if the workload had high change rate as the job must always process the same amount of data. If it's not the case, I'd pay attention to the reported "bottleneck", more information can be found here.

One more useful resource is the best practices guide, you can find best practices for SAP HANA on this page.

Also, please open a support case, upload the support logs package and share the case ID with us. We cannot troubleshoot technical issues over the forum posts effectively, as debug logs analysis is usually required.

Thanks!

[AdrianDiaz]

we have 1 VM that has SUSE installed and we work with SAP HANA, we want to ensure business continuity and therefore a DRP scheme was purchased for this purpose.
We have an additional server identical to the one that has mounted the vmware host and the VM and Veeam was purchased to make backups of the VM and to recover in case of a mishap.

The VM that contains SAP HANA runs its DB and instance backups, to this VM was added a couple of years ago due to lack of storage issues a 1tb ISCSI and when the provider began to configure Veeam showed us an error of lack of space in Datastore1 so it did not run the backups, Due to the lack of support from the provider we started to take the project ourselves, we added an additional 500gb to the ISCSI and the backups ran without any problem, we tested the replication and it was done without any problem. We started the tests to validate the SAP connection, remote desktop and to be able to access the VM, the tests failed and we could not access SAP or the remote desktop, we changed the configuration of the HOST that contains the replication since the IP's of both HOST were in different VLAN's and here began the problems with the replications, since it shows this error:

Processing BF_SUSE12SP4_SAP93PL10_PRODUCTIVE Error: Failed to open VDDK disk [[datastore1] BF_SUSE12SP4_SAP93PL10_PRODUCTIVE_replica_1/BF_SUSE12SP4_SAP93PL10_PRODUCTIVE_3.vmdk] ( is read-only mode - [false] ) Logon attempt with parameters [VC/ESX: [172.16.20.200];Port: 902;Login: [root];VMX Spec: [moref=5];Snapshot mor: [5-snapshot-1];Transports: [nbd];Read Only: [false]] failed because of the following errors: Failed to download disk 'VDDK:vddk://[datastore1] BF_SUSE12SP4_SAP93PL10_PRODUCTIVE_replica_1/B
Replica location '172.16.20.200:datastore1' is getting low on free disk space (0 B left out of 1.7 TB).

Previously it didn't show this and it all started when I changed the IP of the backup HOST.

[Andreas Neufert]

If I understand it correctly, you are using Veeams VMware VM Replication.
You gave the VMware ESXi host a new IP address and now the Veeam Jobs are failing.

Did you have added the ESXi host by Vcenter to Veeam or did you registered the ESXi host directly?

Anyway, make sure that the Veeam Server and Proxy Server (all) can ping the ESXi host and that DNS is working correctly.

Then go to Backup Infrastructure - Manages Servers and right click on the vCenter/Esxi host and go to edit/properties and run through the wizard again.
Likely the ESXi host ceritifcate has changes and therefore you can not access the data. The wizard run will address this and you will be asked about accepting the new certificate.

[AdrianDiaz]

Yes, as I said, we are starting to work with Veeam, we configured the backup and replication jobs and we had several problems, but we were able to get both to work.
We modified the IP of the Vmware host and shortly after we configured as you said, Vcenter with my 2 servers (productive and backup).
Vcenter with my 2 servers (productive and backup).

As you said when adding Vcenter to the Veeam configuration gave me some problems and could be configured again as you said.
The error that marks me started to do just after changing the IP and before Vcenter was configured.

I am trying to submit a support ticket but I have problems with my license because my provider associated the license with another user and I can't submit the ticket at the moment.

[AdrianDiaz]

Processing BF_SUSE12SP4_SAP93PL10_PRODUCTIVE Error: Failed to open VDDK disk [[datastore1] BF_SUSE12SP4_SAP93PL10_PRODUCTIVE_replica/BF_SUSE12SP4_SAP93PL10_PRODUCTIVE_3.vmdk] ( is read-only mode - [false] ) Logon attempt with parameters [VC/ESX: [172.16.20.200];Port: 902;Login: [root];VMX Spec: [moref=6];Snapshot mor: [6-snapshot-1];Transports: [nbd];Read Only: [false]] failed because of the following errors: Failed to download disk 'VDDK:vddk://[datastore1] BF_SUSE12SP4_SAP93PL10_PRODUCTIVE_replica/BF_SU

[Andreas Neufert]

Support need to look at the details. The error message just say that we can not access the file thorugh the ESXi host with IP 172.16.20.200 is this the correct IP of the host?
Can you PING it from all Veeam Servers? Any Firewall in between that could block port TCP 902 and port TCP 443?

[AdrianDiaz]

the IP 172.16.20.200 is the IP of the Backup HOST, we have a firewall but we have not blocked those ports, as I said, previously had the productive HOST with an IP 172.16.20.27 and the IP of the SAP VM is 172.16.20.162, when the new backup HOST and Veeam was added, the HOST was configured in a different IP 172.16.10.21 and the backups and replication worked "correctly". When testing to validate if using replication could connect to SAP and see the info transparently in case of failure of the main server, but could see the VM and OS info entering from the HOST, but could not enter by remote desktop to the IP 172.16.20.162 and when entering the SAP could not find it. 162 and when entering to the SAP it could not find the SLD with the same IP (172.16.20.162. for that reason it was decided to change the HOST of IP and to place the IP 172.16.20.200, supposing that this was the reason why we could not enter to the VM from remote desktop nor from SAP. with this change of IP it was with what the backups nor the replication no longer worked.

[Andreas Neufert]

I think it is time that someone need to look into your environment remotely and see what is going on. Maybe just a small thing but it is really hard to follow here.
Maybe a network diagram would help.
Try maybe Veeam support if they can help you with this infrastructure issue?

[Andreas Neufert]

Dump question...

you say you have two hosts now in the same Subnet 172.16.20.x and you have a firewall in between?

This is usually not working as there could be no routing between systems of the same subnet.

Subnet 172.16.20.x server/hosts ... <=> Firewall (Router) <=> Subnet 172.16.20.x server/hosts
this would not work that way. Only if you use the same Switch for both subnets without the firewall/router in between.

[AdrianDiaz]

Hi. Yes, I have the 2 Host in the same VLAN (20) and in the middle of both is the firewall (Fortinet). apparently the backups and replications are working, only at the end is showing me almost at the end of the job that the HOST runs out of space, the final problem I have is that the transfer speed is not equal to the read speed so it takes a long time to generate the backup, it takes a long time to read the ISCSI that I have in the productive HOST.