Failover setup (2.7)

Open

Configuration Article | CA-20230509-MF-01 VDG Sense | Setup | Failover Open

Introduction

Sense failover functionality has been redesigned in version 2.7 and is more configurable than in older versions.

This article explains the function of the Failover tab. For more information about Failover please read this article: *How failover works .

The Failover tab holds configuration for the following items:

• The servers which need to be monitored

• The clients which need to be informed when a server has been taken over

• Configuration for Failover functions

• Status information

Limitations

• A Failover server can take over 1 server, this means if the failover server has already taken over a failing server and another server fails, this server cannot be taken over

• Only 1 Failover server per Master-Slave configuration (a.k.a. cluster) is supported

Server configuration

The Servers tab contain the monitored servers and corresponding settings for the selected server.

Open

The selected server has the following settings:

• Host

  • The IP address of the monitored server

• Port

  • Predefined, cannot be changed

• Is Master

  • Select if the server is the management server

• Cluster name

  • For future use, should be the same for all servers and clients

• Username / password

  • Credentials (Admin rights) to login to server

• Database config

  • The settings in this menu are used by the Sense failover service to be able to connect to the original database, after a restore. It will perform a merge of the failover database into the original database. After merging, the failover database will be cleared again.

  • In case the selected server is a master server, you should enter the database details in this field. Use the physical ip-address of the database. In normal master-slave situation, the database is running on the management server, so the physical IP address of the management server should be entered.

  • In case the server is a slave server, the database settings will be skipped. It is best to leave all values in the database config menu empty.

• Automatically take over server

  • If enabled the Failover server will automatically take over a failing server that is being monitored

• Take over

  • Manually initiate the take over of the selected server

• Automatically restore server

  • If enabled the Failover server will automatically restore the server it has taken over when it detects the server is up & running again

• Restore

  • Manually initiate the restore when this server is taken over. Can only be done if the failed server is now up & running again. After restore the Failover server will be reset to default state.

• Backup Button '+”

  • To create a new settings backup of the selected server press the + button. During take over of a failing server the last settings backup from that failing server will be used. Automatic back-ups can be configured in the Options menu, but when adding a server for the first time, a manual back-up is required

To add a server to the monitored server list click the ‘+’-button.

Open

• Host

  • The IP address of the monitored server

• Port

  • Predefined, cannot be changed

• Username / password

  • Credentials (Admin rights) to login to server

Client configuration

The Clients tab contains the list of Sense clients which need to be informed when a failing server has been taken over. This is required because the Sense clients are logging in to all servers within a cluster (master/slave setup). This is done based on server IP address. So if one of the servers in the cluster is failing and taken over by the fail over server the Sense clients need to login to the IP address of the failover server instead of that of the failing server.

If clients are not in this list and a server fails, the client will show ‘Server connection lost’ as it is unaware that the failing server has been taken over by the fail over server. It was not notified about the take over and thus it keeps trying to login to the failing server.

Open

To add a client to the client list click the ‘+’-button.

Open

• Host

  • The IP address of the monitored client

• Port

  • Predefined, cannot be changed

• Cluster name

  • For future use, should be the same for all servers and clients

Settings

Open

• Backup Directory

  • The path of the folder where the fail over server will save the monitored servers settings backup. This can be automated. See “Daily backup time”-setting below.

• Store Path

  • The path of the folder where all the video recordings are stored during time that fail over server has taken over failing server.

NOTE: These recordings ‘stay’ on the fail over server. Even after restore when failing server is up & running again.

• Server to Failover Address

  • The IP address of the network adapter of the failover server. This is used in a failover situation for the other servers to communicate with the failover server.

• Client to Failover Address

  • The IP address of the network adapter for the client of the failover server. This is used in a failover situation for the clients to communicate with the failover server.

• Daily Backup

  • Select to enable automatic settings backup of all monitored servers. This will be done once per day.

• Daily backup time

  • Time when the automatic settings backup will be done. IMPORTANT: This is UTC time.

• Failover Database Config

  • The database (Postgres) credentials of the failover server. The database settings on the failover server are not necessary the same as the database settings on the management (master) server.

  • Enter the details of the database configuration on the failover server itself. Use the physical IP address of the failover server. During take over of a management (master) server all the connected slaves within the cluster will be informed to use the failover database configuration.

Timeouts

In the Timeouts subtab, you can set the timeout values of various failed states.

Open

Status

The status tab shows a log of the Failover and which servers are being monitored.

Open

Storage

As explained before, all the video data that is being recording during the time the fail over server has taken over a failing server, stays on the fail over server. It is NOT synched back to the failing server after it is up & running again.

Therefor you must configure on each monitored server within the cluster where it can find its recorded video data on the fail over server. This must be done on the monitored server itself via the Servers → Advanced tab.

There are three settings to configure.

Open

• Failover storepath

  • The location on the failover server where the video data is stored.

  • Correct format is: \\<Failover server IP address>\<monitored server IP address>

  • E.g. \\10.79.11.16\10.79.11.14

NOTE: The video data access is done via Windows file share mechanism. Therefor you must also configure the Windows username and password of the failover server.

• Failover OS username

  • The Operating System (Windows) username of the Failover server.

• Failover OS password

  • The Operating System )Windows) password of the Failover server.