- This topic is empty.
2 March 2022 at 13 h 58 min #11486olivier covasParticipant
Time Navigator has always respected the backup rules provided by VMware by using the features provided in the VMware VDDK libraries.
All our HV VMware backups have always been based on quiesce snapshots to guarantee consistent backups and restores to our customers.
I will describe the functional diagram of a Quiesce backup performed by Time Navigator:
– When a HV VMware backup is triggered, a secured request (SHA1 certificate from the vCenter) is done to the vCenter to determine the list and characteristics of each VM from the VMware datacenter to be backed up.
– Through the VMware VDDK, we submit to the vCenter a backup request for one or more VMs.
This request can be parallelized to simultaneously backup multiple VMs in SAN or VSAN environments. Backup of VM(s) in NSX environments is now supported.
– The backup request is delegated by the vCenter to the ESX managing the VM to be backed up.
– The ESX, through the VMware Tools installed in the Windows VM, lists the VSS writers’ plugins to define which applications are installed in the VM (Active Directory, MS SQL, Exchange, Sharepoint, and any other applications with Writer VSS).
>This is like the Microsoft command: vssadmin list writers.
– After this check, the ESX, through the VMware Tools, performs a Quiesce snapshot by requesting all the VSS writers installed into the Microsoft VM to be backed up.
If one of the VSS Writers is not functional, the VMware Quiesce will not be performed successfully, so the Time Navigator backup will return an error on this VM.
During the Quiesce of a Windows 2019 VM, with SQL Server installed, we could see in the Windows Event Viewer system logs, an error of type event id: 513
Log Name: Application
Event ID: 513
Task Category: none
Access is denied.
The correction of this error, using the Microsoft technote below, allowed to correct the VMware quiesce error:
– When the Quiesce snapshot is successfully completed, the vCenter attempts to present the Datastore (where the snapshot is located) to the HV VMware proxy via the SAN if the proxy has a SAN attachment and zoning has been done.
– If the HV VMware proxy is a VM, the hotadd mode is used (similar to the SAN inside the VMware infrastructure).
– If the HV VMware proxy does not have a SAN attachment (no HBA controller) or if the SAN attachment is lost, the ESX will transfer data over the network (NBD mode) via TCP port 902 to the HV VMware proxy.
If the VMware infrastructure is in VSAN mode, we can use a HV VMware multi-proxy mode to use the hotadd mode for all backups (according to the VMware prerequisites).
TINA_VCB_VIRTUAL_HOST tina variable
– When Time Navigator has read all the data (SAN or hotadd mode) or received all the data (NBD mode), it requests the deletion of the snapshot from vCenter.
- You must be logged in to reply to this topic.
Sujets / Topics
- Miria 2022 (3.15) is available
- Lina Linux agent space issue Tracker.db
- How to generate an AER in MIRIA or ADA
- Lina is not impacted by the SpringShell vulnerability CVE-2022-22965
- Tina is not impacted by the SpringShell vulnerability CVE-2022-22965
- Miria is not impacted by the SpringShell vulnerability CVE-2022-22965
- New Tech Paper on M365
- Advisory ID : MIRIA-2022-0001
- Do not use macOS UI 3.14 !
- Miria is not impacted by vulnerability CVE-2021-40438