NCM Config Archive Replication?

Question

I've been assigned a task to replicate the NCM backup archive between pollers for redundancy and DR. What's the best way to do this? I'm more Python-oriented and have been using the OrionSDK library with the API to make calls to the NCM.ConfigArchive API, but here are the main questions with this approach:

* The configurations are not user-friendly. They have a ConfigID and NodeID that that provide much help. They're both just long text strings with letters and numbers, and I'd like to replicate the files in a way that makes sense. IE: Device-name_Date_Config.txt. What is the best way to decipher the NodeID to an actual Node hostname?
* What's the best way to replicate only the most current configuration the database has?
* What's the best way to replicate all of this to another poller running Windows Server 2012?

mesverrum · Answer

You don't have to keep the config archives locally on the pollers. Just set them to all write to a UNC path for a file server that already has redundancy and replication set up. Solves the problem of them being scattered, and piggy backs off the fact that DR file replication should be a solved problem already so you don't have to try to cook up some custom one-off that nobody will remember how it works 5 years from now.

cnorborg · Answer

Oh, almost forgot.   Where you set up the config archive, its a path that you archive the configs to, this could even be on another server or fileshare.

Additionally, there is a place to specify a template for the naming of the config files.   Initially its:

${NodeGroup}\${Caption}\${Date}\${Caption}-${ConfigType}.Config

So, it saves the file into a directory based on their group, then their caption or hostname (so all configs for a given box are in a single folder), then a folder by date, then the filename is the hostname with a dash then the config type (running or startup) followed by ".config".

From what you said you would probably want it to be:

${Caption}_${Date}_Config.txt

Although there is no distinguishing between running and startup configs in this manner.   But you have the right to name it what you want!

cnorborg · Answer

Ok, when you refer to "ConfigID" and "NodeID", you sound as if you are referring to the database of the Config Archives, which is contained on the database server.   While you might have quite a few pollers, you only have a single database server.   So there is no way to replicate this data to the pollers simply because they don't have a database on them.

If you're referring to the "Config Archive Folders", where NCM additionally stores a plain text copy of the config.   That gets more tricky.   If you want to write a script to do this based off the database using the API, I think you're taking and awful lot of effort for something you probably don't have to do.   The text versions of the configs really aren't that big.   It's quite simple to set up windows to replicate the folders you put these configs in somewhere else.   Personally I'd put it on a file server that is getting backed up rather than another poller.    Easiest would be to copy all the files on a regular basis to a separate folder for each poller, but you could probably with little or no effort even combine these together into one folder.