For most maintenance you need to do on physical hardware in the Collage environment, you simply make your changes (e.g, add memory) and move the node from the Collage maintenance pool back to the Collage discovered pool. Collage then makes the node available for use. However, if you are replacing the motherboard or the primary boot NIC on the node, you need to follow a slightly different procedure. The reason is, Collage may think that it has already discovered that node and won't re-discover it to bring it into the free pool. So, if you've got a node with an unhappy mobo or you are replacing the primary boot NIC, here's what you need to do.

1. Log in to the Collage Controller.
2. Find the node in question and click the checkbox next to it.
3. Select Node Actions > Delete Node:
   
   This removes all knowledge of the node from the Collage database.
4. Replace the motherboard.
5. Power the node on again.
6. Collage discovers and inventories the node as if it were a new node being added to the environment.
   

Remember, for most maintenance, you don't need to go through the step of deleting the node from Collage. Just make your changes and move the node back into the discovered pool. If replacing a motherboard or primary boot NIC, however, you need to also delete the node from Collage.

Labels: Collage maintenance operations motherboard primary boot NIC

By default, Collage logging information is captured in /opt/cassatt/logs/collage-core.log. However, many system administrators like to streamline system management by using syslog to consolidate log info in one location. If you would like to redirect Collage errors, warnings, or actions to your system log file, use this procedure:

1. Edit /etc/sysconfig/syslog and add -r to the SYSLOGD_OPTIONS. For example:
   SYSLOGD_OPTIONS="-m 0 -r"
2. Restart syslog.
3. Edit /cassatt/etc/log4j.properties and add syslog to the rootCategory. When done, the rootCategory entry should look like this:
   log4j.rootCategory=warn, ccore, syslog
   
   While you're editing the log4j.properties file, note that the default syslog Facility setting (the field that is meant to indicate where in the system the message originated) is local0. You can change that according to your site policy if you need to.
   
   
4. Restart the Collage service using the following command:
   /opt/cassatt/bin/cccoreservice restart

That's all there is to it. Many enterprise sites will likely run some kind of central syslog server. If you work in this kind of environment, or if you just want the convenience of integrating Collage log information into a common logging location, you can use this procedure to redirect Collage events to your system log.

Labels: Collage logs syslog log4j.properties

In a dual-control node configuration, there are operations procedures (for example, backup and restore) where you need to know which node is the active and which is the standby control node. The instructions to do that are described in the relevant documentation, but I thought it might be good to capture it in this forum as a general reminder. The basic step is to log in to a control node as the root user and use the cccoreservice command with the status option:
/opt/cassatt/bin/cccoreservice status
The output will look something like this:

As you can see from the annotation in the sample output, the key is to find which node (identified by its IP address) has the collage-core service started. That's the one currently providing Collage services, making it the active control node.

Labels: collage "control nodes"

The SSL certificate (the bit of code that guarantees secure web transactions) that is included in Collage expires on May 2, 2007. What does that mean to you? When you log in to the Collage Controller, you will see a popup alert that indicates the certificate has expired. However, you can simply accept the expired certificate and continue. The alert can be a little annoying, but Collage will continue to operate normally.

The next generally available release of Collage will include an updated SSL certificate. Until then, if you want to avoid the popup alert, you can use the following procedure to update existing Collage 3.4—4.0 systems. Here's what you do:

1. Go to the Cassatt customer FTP site: ftp.cassatt.com.
2. Log in as anonymous.
3. Get the SSL certificate file from the following location:
   outbound/customer/keystore.zip
4. On all control nodes, extract the new SSL certificate file into the following directory:
   
   /opt/cassatt/server/
   
   You will need root access. Also, it's okay to overwrite the existing file.
5. Restart Collage. Use one of the following procedures, depending on whether you are using a single-control node or dual-control node implementation:

   If using...	Then follow these steps...
   Single-control node system	1. Log in to the control node as root. 2. Use the following cccoreservice command to restart the Collage service: /opt/cassatt/bin/cccoreservice restart
   Dual-control node system	1. Use the following cccoreservice command to determine the active control node: /opt/cassatt/bin/cccoreservice status In the output, the node with Status equal to "started" is the active control node. 2. Log in as root to the active control node. 3. Use the following cccoreservice command to restart the Collage service: /opt/cassatt/bin/cccoreservice restart

   
   Note that you can use the same cccoreservice command for either single- or dual-control node systems. Previous versions of Collage required different commands.

That's all there is to it. Once you update the keystore file, the popup alerts will go away.

Labels: Collage SSL certification "control node"