Overview

The following article describes the procedure of updating Zevenet 6 Cluster or higher version in order to obtain the minimum downtime during the update procedure. According to the services configured, you can experience a minimum of downtime or no downtime at all.

If no cluster is configured, the downtime could be incremented for one minute in the worst case.

Note that if you are currently running an older Zevenet version than 6 then you should review if your cluster is ready to be migrated according to the article Configure Zevenet 5 Cluster.

Environment

In order to understand better the update procedure with minimum downtime, the following environment has been built:

Requirements from EE 6

Connection to the internet is required to check in our repositories pending packages, in case your system is not able to connect to the internet please check upgrades with the offline procedure,

Zevenet 6 Offline update packages

Update Procedure

Since ZEVENET 6.0.0 the updates are done with the tool checkupgrades.

Before starting, please execute a snapshot of both nodes of the cluster, just to revert changes if needed.

0. Go to the node with the MASTER role and stop the zeninotify process, this process which is responsible of replicate any config change, also lock any sync request from the backup node as indicated below.

[master]root@zva5n1:# kill -9 `ps -ef | grep zeninotify | grep enterprise.bin | awk {'print $2'}`

In the master node, block any ssh connection request from the BACKUP For this open the file /etc/hosts.deny and enter the following line at the end of the file. Please change 192.168.100.25 by your current BACKUP IP. Save changes and exit. This action will block in the MASTER node any try of connection to ssh service from BACKUP.

(in the current MASTER, file /etc/hosts.deny)
sshd: 192.168.100.25

1. Go to the BACKUP node (following the current example zva5n2), and apply the update as described in the instructions of the update package, or execute checkupgrades tool if you run Zevenet 6.0.0 or later:

[backup]root@zlb:# checkupgrades
[backup]root@zlb:# checkupgrades -i

2. Once the update process concludes verify that it finished without errors. If some error is detected in the installation process please open a support ticket and attach the output of the installation process to it.

3. If there aren’t errors during the update of the BACKUP node (zva5n2), go to the MASTER node (zva5n1) and apply now the update in this server as indicated in step 1. Once the update is done, the service will be restarted in order to apply the changes, hence during this process the service will switch automatically and the current BACKUP node will become the new MASTER.

4. If the update is done without any error then proceed to disable the already configured sshd block in /etc/hosts.deny deleting the line entered in step 0 in the old MASTER node zna5n1.

5. If the fail-back is enabled then the MASTER role will be moved automatically to this node again. If this is not the case, the MASTER role will be kept in the second node (zva5n2).

6. In order to force the MASTER role to node zva5n1 then you only have to enable the maintenance mode in node zva5n2, and the MASTER role will be taken by the node zva5n1 automatically as just to before the update. Once

Once the migration process finishes please ensure to disable the maintenance mode of both nodes and confirm which node is MASTER and which node is BACKUP in order to continue managing the system through MASTER. Ensure that the temporary changes in the file /etc/hosts.deny are deleted. Finally you can check the cluster service status since the view SYSTEM > Cluster.

Enjoy your just updated Zevenet without downtime!