Tembria Blog

We’re seeing an issue in v5.5.3 where after a clean install the reports may not display after they are built.  The problem seems to be related to the fact that some folders are not being created until the first periodic save, which happens after ten minutes, and when the report builder kicks in it exits because it can’t find the folder it is supposed to write to.  Happily the periodic save literally saves the day and after the monitoring service has been running for ten minutes all will be well.  We have fixed this in our internal builds and the fix will be included in v5.5.4 which will be released in about four weeks.

Posted by Don

In Tembria Server Monitor v5.5.3, which is currently due on or around January 21st, we have made a few updates to the administration history.  Since this is a feature that is often overlooked, even by longtime Tembria users, we thought we would give it some coverage here on the blog.

The administration history shows a comprehensive list of all changes that were made to the monitoring configuration.  It’s a great way to keep track of what has been modified and when, especially in environments where there are multiple network admins using Tembria Server Monitor.

A couple of issues with the administration history were reported recently and we have fixes for them coming in v5.5.3.  First, the filters at the bottom of the page show incorrect information.  Instead of showing user names and categories they are showing incorrect values.  That’s been fixed.  Also we found that there were a couple of actions in the user interface that were not being recorded in the administration history.  We fixed those too.

Along the way we also updated our Active Notifications so that the actions taken through them are logged in the administration history as well.  Active Notifications are another feature that deserves more attention.  With Active Notifications you can reply to our e-mail notifications with instructions for the monitoring engine such  “tembria pause device 5 minutes”.

Posted by Dave

For the event monitors that need to authenticate to a remote device, the settings section includes a number of different options and some are new as of Tembria Server Monitor v5.5.  Using them correctly can help to increase monitoring throughput, generally letting you monitor more things more frequently.

The Auth Type option lets you select how Tembria Server Monitor will authentication to the remote machine.  With “Domain/Local Authentication” we authenticate to the domain and then reach out to the remote machine in the security context granted by the domain controller.  That means that the remote machine needs to be a member of the domain as well, or a domain trusted by the one we authenticated to.  This option has the advantage of being relatively fast.

With the “Remote Authentication” option we reach out to the remote machine and then authenticate to it.  This has the advantage that the remote machine can be in another domain or in a workgroup and authentication will still work.  The downside is that it can take a while for the remote authentication procedure to complete.  It not usually an issue but if you are monitoring hundreds of machines, the time can add up.

The remaining authentication options are used to select the user name password that will be used to authenticate.

The first option, called “Use the monitoring service account”, tells the event monitor to use whatever account the monitoring service is logged into.  This option is disabled by default because our installer puts the service in the LocalSystem account, where most services run.  To enable it, go to the Services applet in the Administrative Tools section of the Window Control Panel.   In the properties for the Tembria Server Monitor Service you can use the Log On tab to select the account that our monitoring service runs in.  Choose an account with administrative privileges for the majority of the machines to be monitored and then restart the monitoring service.  Back in the Tembria user interface you’ll see the option is now available.  This option has the highest performance of all because the monitoring engine is automatically authenticated and the event monitors can simply connect to the remote machines and get their jobs done.

The next option to “Use and authentication profile” becomes available after you have defined one or more authentication profiles.  An authentication profile lets you give a name to a user name / password pair and then you can select those credentials with the profile name.  If you need to change the user name or password at a later date, you just change them in the authentication profile and all the event monitors pick up the change automatically.

The last option “Use these credentials” allows you to directly specify the user name and password to be used when authenticating.  It’s great for quick and easy tests or monitoring devices that need special credentials but it’s more difficult to manage in the long run.

So to sum up, for optimal monitoring performance use the option to “Use the monitoring service account” and use “Domain/Local Authentication” wherever possible.  Use the other options as necessary in order to connect to devices that need separate credentials.

Posted by Don

We’re very happy to announce the release of Tembria Server Monitor v5.5.2!   Probably the most important update in this release is a fix for an issue that could cause period slowdowns when navigating in the interface.  We tracked this down to a problem in our integrated web server that was introduced in v5.5.0.  With v5.5.2 you’ll see that the user interface is back to its consistent snappy performance, even for large configurations.

A couple of other notable changes include a faster check for CPU usage in the System Health event monitor and a fix for a problem where after moving a device to a new group it might not appear there until you refresh the browser.

As always the complete list of changes can be found in our version history.

Posted by Don

We have had a few reports of periodic delays when accessing the Tembria Server Monitor v5.5 through the integrated web server, which is the default method.  The problem seems to be specific to certain browser/OS combinations and we’re running some tests now to get a better idea of where the problem lies.  The main symptom is that some times you’ll select an item in the user interface instead of coming back right away it can take a while for the page to refresh, even though the system seems to be idle and there’s nothing else going on.  If you’re affected by this problem and it’s causing frustrations, get in touch with our support guys and they’ll get you temporarily switched to IIS mode, which works around the problem.

Update: The issue described above has been fixed in v5.5.2.

Posted by Dave

We released version 5.5.1 of Tembria Server Monitor today and we encourage you to download and upgrade when you have a chance.  Most of the fixes included in this release are for relatively minor issues but there are some performance enhancements as well that everyone will benefit from.  In particular, graph generation has been overhauled and you’ll especially notice that custom dashboards with lots of graphs will display much more quickly.

A complete list of all the changes can be found in our version history.

Post by Don

With a certain amount of exhaustion we are very pleased to announce the release of Tembria Server Monitor v5.5! This new release includes a wealth of new monitoring features all tucked into our existing user interface.

One of our most technically challenging updates in v5.5 is pretty much invisible but important nonetheless:  We rewrote our monitoring engine to introduce a lot more parallelism and allow it to scale to much larger installations.  We’ve come a long way since v1.0 was released in 2003.  It was a desktop utility that could monitor about 25 machines on the local subnet.  We’re now rolling out deployments for environments with 2500+ devices to monitor and running tests in environments that simulate 5000+ systems under active monitoring.  We modestly believe that pretty good for a product like our Enterprise Edition at a modest price of $1995 USD.

Most of the other changes in v5.5 are more obvious and I’d like to highlight a few of the ones that have been the big hits with our beta testers and we think you’ll greatly benefit from them too.

Our new logbook functionality is not obvious at first glance but most people who start to use it quickly find that it has huge benefit.  We’ll be writing more about this in a separate post but here is a quick overview:  Let’s say Tembria Server Monitor sends you an alert telling you that the H: drive on a file server is approaching capacity.  Naturally, you investigate and you find that a user has dumped a pile of personal data onto a share that is supposed to be only for corporate data.  So you back it up, clear it out and inform the user.

You’ve been alerted to a problem and have taken corrective action.   Tembria Server Monitor v5.5 now lets you close the loop.  You select the event history option to add a log book entry.  Then you enter details about what occurred and what action you took.  Over time you’ll build up a great history of all the administrative actions you took in order to keep things running smoothly.

Other new features in v5.5 include a new System Restart event monitor that alerts you about restart events while distinguishing between planned and unplanned restarts and even providing the reason text entered by administrators.  Our SNMP Trap event monitor implements caching allowing it to better handle floods of traps, which is the way traps tend to arrive of course.  A new SSH/Linux event monitor lets you run any command you choose on remote systems support SSH and includes sophisticated parsing options to let you make sense of command results and generate the appropriate notifications.  New notification types give you more control over when you receive notifications and how many you receive for ongoing error conditions.

And there’s a lot more than that.  It’s all covered in our version history and we encourage you to take peek before upgrading.

We would like to express our thanks to everyone who participated in the beta testing of v5.5.  With your help we were able to track down and fix many issues and your comments and suggestions were very much appreciated and very valuable to us.

Posted by Don

Version 5.5 is probably our most eagerly anticipated release ever and we’re getting a lot of questions about when it will be available.  It’s in the final stages of testing at the moment and we need just a bit more time to address some issues with upgrading from earlier releases and to complete some documentation tasks.  We’re aiming for a release on October 15th and everything looks great for that date.

A note to our beta testers: In a day or two we’ll be posting an update in our pre-release web site.  It will likely be the last update before the official release.

Posted by Don