ServerGraph - A TSM Management and Automation tool

Overview

Servergraph is a useful browser based reporting and automation tool for managing multiple TSM servers. It extracts data from TSM logs and databases and combines these together to give you a composite picture of your TSM estate. It can issue problem alerts, provide trend graphs, automate manual tasks and more. Originally a TSM tool, Servergraph has been recently enhanced to support Netbackup and Legato too. This page concentrates on TSM management.

Installation

Servergraph is reasonably easy to install. It comes supplied with detailed documentation so there is no point in reproducing it here. You have two install options; a normal fully automated install that will update your system files after taking backup copies, or a manual install with a -m switch. With a manual install you are fully in control of what gets updated, but it takes a lot longer and you have to take your own system backups.
Some third party software is required for the install, including MySQL DBMS, Apache web server and Perl. The automatic install will check for these and install them for you, if they do not exist already.
One advantage Servergraph has over other TSM reporting products is that it only needs to be installed on one machine, it does not need agents installed on every TSM server or client.

Features

Servergraph extracts data from the TSM databases and logs and formats that data into its own MySQL database. That minimises the performance overhead on TSM, as data processing and interpretation is carried out away from the TSM databases

ServerGraph Alerting traps all the TSM error messages and picks out the two hundred or so that are important. It can pass these to Tivoli TEC or other messaging products as SNPE alerts.

It is possible to tailor the Servergraph front logon screen, but the default provides lots of useful information, including a summary of the results from the previous day's backups, a 30 day backup history, a summary of the state of the databases and logs and a summary of the state of the tape volume and drives. This lets you get an instant impression of the state of the backups and backup infrastructure on one screen. You can then drill down into most of these summaries and get detail, and extract that detail as Excel or csv files

You can also expand the left hand menu to get details about the whole enterprise, and from individual servers. Expand the 'enterprise' menu then the 'nodes' menu and select 'top 50 storage abusers'. This will list fifty nodes sorted by their 'hog factor', which is the ratio of primary to local storage. I've found this report to be very useful to identify nodes where the retention period is incorrect, and so is keeping backups for far too long.
While the picture above just shows TSM servers, if you run Legato and / or NetBackup, ServerGraph gives you a composite picture of how all three are working from one screen

From the same Enterprise menu, select 'Long Term Trends' then highlight '90 days' and you see a report like this

The 'Predictions' option shown above displays charts that illustrate how the capacity of databases, logs and storage pools is changing and when they will run out. This is useful for both Capacity and Budget planning.

Tips

A particularly useful graph is one which shows each nodes behavior over 24 hours. This data can be used to detect nodes suffering from speed problems due to incorrect network card settings.

Some of the ServerGraph pages are wider than a screen and you have to scroll right to see the edge. This can be a bit annoying, but you can hide the Left side navigation menu by going into Main > Hide tree then you should not need to scroll. To see the navigation tree again, use Main > Show Tree.

Function key F5 will take you back to the primary menu from wherever you are in ServerGraph.

ServerGraph communicates with the TSM servers on port 1500 by default. You must ensure that this port is open and accessible through your firewalls.

ServerGraph 3.5 New Features

The following is a short summary of some of the changes and new features in ServerGraph 3.2

User Views allows you to group nodes and users together, so you can; restrict some user's views to a subset of nodes; group nodes by business unit, geographical area or any other grouping that makes sense to you.

The health check graphs now display the volume count in a pool.

The library volumes table has been enhanced to display volume utilisation, error rates, scratch status and DRM status on one screen.

A help box is available to explain the meanings of the column names on the node data table.

back to top

Copyright © Lascon Storage Ltd. 2000 to present date. By entering and using this site, you accept the conditions and limitations of use