resilient to fault tolerance

Software-based VM-centric and flash-friendly VM storage + free version

Moderators: anton (staff), art (staff), Max (staff), Anatoly (staff)

Post Reply
alman64
Posts: 4
Joined: Fri Sep 10, 2021 6:24 am

Fri Sep 10, 2021 7:19 am

Hi, does this system resist the fault of the VM with the SAN software or a fault of host with this VM? I'm doing some tests and if I turn off the VM with a clean shutdown everything is ok, otherwise the storage even if synchronized in two ways is no longer usable
yaroslav (staff)
Staff
Posts: 2279
Joined: Mon Nov 18, 2019 11:11 am

Fri Sep 10, 2021 7:32 am

Welcome to StarWind Forum.
Software tolerates the failure of a host with the VM.
May I have a detailed description of your environment?
Are you sure that HA devices are synchronized between hosts (i.e., you created HA devices, not img files)?
See the recommended shutdown procedure https://knowledgebase.starwindsoftware. ... installed/ under To restart the servers for installing software updates, hardware changes, etc:. Make sure that the prerequisites for restart are met before the shutdown.
alman64
Posts: 4
Joined: Fri Sep 10, 2021 6:24 am

Sat Sep 11, 2021 10:05 am

Thanks for the reply yaroslav! The test environment where I am testing the product is composed as follows:
2 ESX 6.7 U3 with two disks in RAID1 of 144 GB and 40 GB of RAM
On these two hosts I have installed and configured the starwind appliances as per guide:
- 4 Network Interfaces:
1) Management
2) iSCSI and HB
3) iSCSI Sync 1
4) iSCSI Sync 2

I then added a 50 GB disk on the two VMs and created an HA device with the console using the replication manager.

If you need more details, please ask.

I am testing the product to see if it is resilient to the voltage drops that have occurred in the CED and consequently without the possibility of doing a clean shutdown.

If I do the latter there are no problems but if I turn off the machine or the host directly the test machine stops working and in some cases the VMs have also been corrupted, both the test and the one where the StarWind software runs.

Another question, with the free version is the Console necessary for the correct functioning of the vSAN or is it only needed to manage the environment more easily?

Thanks
alman64
Posts: 4
Joined: Fri Sep 10, 2021 6:24 am

Sat Sep 11, 2021 10:17 am

Trying again now a simple clean shutdown of one of the two appliances with starwind now no longer starts and presents the emergency screen
"Welcome to emergency mode! After logging in, type" journalctl -xb ... ".
This he had never done before
yaroslav (staff)
Staff
Posts: 2279
Joined: Mon Nov 18, 2019 11:11 am

Sat Sep 11, 2021 10:36 am

There must be a bad mount you need to check fstab.
If you know root password type it, otherwise, try rescue media to rescue the OS, or redeploy the VM using old disks.
Do you have any data on those VMs?
alman64
Posts: 4
Joined: Fri Sep 10, 2021 6:24 am

Sat Sep 11, 2021 10:54 am

Yes, in fact editing fstab there were two UUID lines at the bottom. Commenting on the first, the system VM with starwind has restarted. The VM test is simple machine without data.

Regarding the previous post, can you answer me as soon as you can (shutdown not clean for voltage drops and console) ? Thanks
yaroslav (staff)
Staff
Posts: 2279
Joined: Mon Nov 18, 2019 11:11 am

Sat Sep 11, 2021 5:29 pm

Hi,

Are you sure the storage from both VMs is connected to both hosts?
Management console helps managing the environment easier.
Most of the every-day tasks can be done in StarWindX scripts. Install the Management Console with Integrated Library option enabled.
Post Reply