I'd like the ability to create custom views under node management. I'd also like to be able to sort by clicking on the column header. I have found that many of our workstations come up as unknown. Trying to scroll through 1000s of workstations to find certain class of workstation is a tedious tasks as it overload the scroll window easily. I'd all like to easily be able to select groups of workstations and unmanage them indefinately. Along that same line it would be nice if we could define group classes that have common settings. For instance workstations come online and go offline frequently in our environment. I want them in NPM, but I don't want NPM to manage them unless I have a specific need to. Lets say I also want servers at certain locations to only be polled every 30 minutes, but servers at my primary sites every 10 minutes, my local servers every 5 minutes. Right now If I want to make those settings I have to go in an make changes to 200 different objects.
I would love to see a small icon on the summary page telling me that an upgrade is available. I have so many applications to keep up to date that any little bit helps.
It would be very helpful to make multiple custom mibs showin one chart. Example: QoS groups and then the DSCP within the group and then the usage of each.
A second example: being able to chart a custome mib for services on a device such as a windows server looking for service 443 or 80 or some such. If the service is not "alive", the chart reflects that and an alert is sent. (Windows is a good platform to start on.
Make custom MIBs much easier to use on the Node page for the WEB display.
Better Netflow display as well.
I would like to have the ability to add a link for the "Explanation" of a MIB entry to the web console.
For example,
interval.
Graph enhancements (especially for custom pollers)- set x axis range and step interval. It can be difficult to interpret graphical information when only the polled values are displayed rather than an allowed range. Allow an option to dynamically expand a graph beyond the given range if polled values exceed the configured range. - ability to include more than one poller in a graph- ability to include a legend with poller information: min/avg/max/current/95th percentil/etc- ability to define reference lines: example - solid line at interface speed, warning/critical threshold, etc- currently can bulk modify custom poller graphs, want to be able to override template for single graph- ability to label units on vertical access- modifying date range intelligently modifies sample interval to only include acceptable ranges- ability to monitor "real-time" (specify shorter interval) poller output without requiring activex or the engineers toolset.I come to Orion from Cacti and the graphing in Orion leaves much to be desired in comparison. The graphs are more difficult to interpret, more difficult to customize, some customizations are not possible, do not allow logical groupings of multiple pollers in a single graph, and quite frankly don't have the same level of finish as the ones in cacti.
Alerting enhancements- integrate syslog/trap alerts so that node information is available- ability to include trap details as a variable in alert messages, instead of entire message (sms limits characters and pertinent info cannot always be displayed)- allow a trap to be cleared by a corresponding trap within a given interval (i.e. I don't need a page if my ups drops to battery power for 1s, but would like a page if service is not restored within a longer window. The ups will send a trap when it drops to battery power and when it returns to main power. I'd like the ability to clear the alert before any action is taken if that second alert is seen in time).- integrate syslog/trap alerts so that they can insert NPM Events into the web interface- Baseline alerts: compare to last poll, last x polls, last hour at this time, yesterday at this time, last week at this time, arbitrary time. % change, amount of change. - Ensure that modifications to alerts take effect immediately... that random delay thing is a real pita- discarded traps are not processed for following alerts- ability to create an alert quickly and easily for a class of devices or a single device by clicking on any given graph.- (Added 11/13) Allow the creation of a full set of alert actions after x time for alert escalation. This could be approximated by triggering an alert if a condition exists for a short period of time and then delaying other actions, and creating a second alert that waits a longer period of time before it is triggered.
Ability to schedule recurring windows where a device is not managed: devices shut down on weekends, rebooted at a regular interval, etc.
Tabbed views for departmental access. Each dept has similar devices and would like to quickly and easily group them, but still have access to all departments (if allowed)
Suppress events - we have wireless backhauls that report an interface speed change when they change modulation. This is constantly dumped into the event log for NPM. I expect this and don't want to see those messages since they only clutter the display.
Force basic categories: for example I have one or two devices that despite having the same hardware and firmware and configuration are not properly categorized with the correct vendor. I would like to be able to force this in those cases to make up for vendor shortcomings.
Adjust interface utilization percentage graphs to account for the fact that links may not have the same speed in both directions (wireless for example). RX percentage below is actually around 80%, not 40%.
Adding another one: A node menu that is accessible from every page, drop down, pane on one side or another, however it fits. It is time consuming to drill down into a node, then go back to Home, expand the tree, search out your device, click on it, etc... Maybe an intelligent browse to selection that would allow you to enter a node name or IP, etc and orion would intelligently provide suggestions to finish off the provided information to allow you to pull up the correct node - similar to the search suggestions provided in firefox for example. That would allow a quick link to any device from any page in the web interface.
Good suggestions.
Tabbed Views !!.. that's a very good one... There should be a Tabbed View in the same page where we could Select which Group Nodes we would like to see (grouping nodes before with a custom field, of course).. All this, without changing to a complety different web-page.
Maybe both: Vertical (on the left) and Horizontal (on top) for being able to group like this:
On Left: Tabbed by CustomProperty1
On Top: Tabbed by CustomProperty2
--------------------
All this besides changing behavior of The MENU... instead of leaving current "Top 10" Tab.. you could go to "Admin" Tab, make changes.. submit.. and then return to "Top 10" Tab... and so on.. Just an Idea.
Tabbed views in the Node Details and Interface Details pages would also be very nice to have and would eliminate a lot of unwanted scrolling.
I really wanted to setup RSS feeds on our Orion home page but Orion did not provide a way to easily do this. I found a way to do this by adding a custom HTML resource to the page and using the RSS widget from yourminis.com. To do this just go to:http://www.yourminis.com/minis/yourminis/yourminis/mini:rsscontainer Set your widget parameters the way you want (color, size, title, content, etc...) then copy the code unto the custom HTML resource and voila! You now have RSS feeds on your Orion web site.
Another thing I've done is setup a Menu item that the status of all of our IP subnets. I did this by using Engineer's Toolkit's IP Address Management tool. I have it setup to publish all of the subnets to a web page every couple of minutes. Then I added the IP Address management web page URL to an Orion menu item.