Upgrading from Pre-4.3 NorthStar with Analytics
The special upgrade procedure described in this topic is only necessary if you use NorthStar Analytics and you are upgrading NorthStar from a version of NorthStar earlier than Release 4.3. In all other cases, you can install NorthStar using the procedure described in Installing the NorthStar Controller.
This procedure involves uninstalling and reinstalling NorthStar analytics in addition to upgrading the NorthStar application, and is a little different depending on whether you have analytics installed on the same server as the NorthStar application or not.
Export Existing Data from the NorthStar Application Server (Recommended)
This procedure involves running a utility called es_export_import_util.py on the NorthStar application server to export and save your existing data prior to upgrade.
The utility exports data to a file called exportdata.tar.Z in the /opt/northstar/northstar_bundle_x.x.x/db_migration directory. Ensure that this directory has available space equivalent to at least 10% the size of your Elasticsearch database. For example, a 10GB Elasticsearch database would require 1GB in the /opt/northstar/northstar_bundle_x.x.x/db_migration directory. When you begin this procedure, the script will tell you how much memory is required and give you the option to stop the procedure if you do not have enough space to continue.
The amount of time required for the export utility to complete depends on the number of export days and Elasticsearch cache memory/CPU cores you have. You can use the utility’s -l option to reduce the number of export days.
You can use the following command to see the utility’s full set of supported export options:
es_export_import_util.py --help
To export and save your existing data, use the following procedure:
Upgrade Procedure with NorthStar Application and NorthStar Analytics on the Same Server
Use the following procedure:
Upgrade Procedure with NorthStar Application and NorthStar Analytics on Separate Servers
Use the following procedure:
Update the Netflow Aggregation Setting
The netflowd process analyzes traffic from the router and displays it in the Demands tab in the network information table. By default, Netflow aggregates traffic by PE, but for some applications (such as EPE), you would want the traffic aggregated by prefix. The Netflow aggregate by prefix setting controls how traffic is aggregated. If the parameter is enabled, Netflow aggregates all traffic from a specific ingress PE router to a specific destination (prefix) within the defined period of time.
Between NorthStar Controller Releases 4.2.0 and 4.3.0, the possible values for the Netflow aggregate_by_prefix parameter changed. If you are upgrading from Release 4.2.0 or earlier, you must modify the setting to reflect a valid value for your post-4.2.0 release. See Table 1
NorthStar Releases Earlier than 4.2.0 |
NorthStar Releases 4.3.0 and Later |
|
---|---|---|
Netflow aggregate by prefix possible values |
Note:
NorthStar Releases 4.2.0 and earlier: netflow_aggregate_by_prefix parameter maintained in the northstar.cfg file. |
Note:
NorthStar Releases 4.3.0 through 6.0.0: modify the netflow_aggregate_by_prefix parameter in the /opt/northstar/data/northstar.cfg file. NorthStar Releases 6.1.0 and later: modify the aggregate-by-prefix
setting using the NorthStar CLI |
Import Existing Data (Recommended)
In this procedure, you run the es_export_import_util.py utility again on the NorthStar application server to import the data you previously exported and saved.