0

I develop a monitoring system that continuously monitors remote hosts and provides a real-time report about the status of some components of this remote host.

For each component, I provide one of these options:

enter image description here

And here is a part of how it looks like:

enter image description here

The system has several components in which the way to check them is as following:

  1. Connect to remote host (through SSH or HTTP request)
  2. Check (poll) the component's status and get a result
  3. Returns the appropriate result (operating normally; Error; Warning etc.)

For instance, assuming I'd like to monitor the usage of memory component in a given remote host and provide an error if users using too much memory, firstly I have to (a) connect to this host, then I have to (b) collect data and (c) returns it.

My dilemma is about situations that you can't provide the data.

  • That could be because you can't connect to host
  • That could be because you can connect to host but it doesn't return a value (for some technical reasons, e.g. a problem in operating system)

My question: Would you provide a red icon of "Error" or gray icon of "check status is unknown" in case you can't provide data about the memory usage? Which one better reflects the situation?

Currently I use red "Error" icon in case that users use too much memory OR I can't provide data, but I'm not sure if the second situation should really provide red "Error icon". What do you think about?

THANK YOU

  • 2
    This isn't directly related to your question, but one suggestion I wanted to make is that you consider minimizing the visual interest when "service is operating normally". Ideally, at any given moment, your UI would show 100% green icons if the service is operating as expected. I'm not sure how valuable a screen full of green icons are, because in this case, no news is great news. As a user I'm probably more concerned with the outliers and problems; which services aren't functioning so I can resolve them, or which are not communicating. – Ben Harrison Aug 23 '16 at 16:25
  • 1
    I was just about to make the point that @BenHarrison made. The default state is "normal," so there should not be an icon for that. Rather indicate only the exceptions for a clearer interface. Especially when the icons for normal and error are so similar (doubly so for the colorblind). – Eric Stoltz Aug 23 '16 at 18:29
  • @Ben you're right. Actually we provide a checkbox (you can't see) that filters out all green icons (if user wants) for the reason you mentioned. – Tamir Gefen Aug 24 '16 at 13:41
1

First of all, get rid of ambiguity

You're displaying an error because your users are consuming too much bandwidth, yet you're showing an arrow down (which communicates less rather than more).

Second, you need to be clear about what's going on, without a doubt. You can't display error messages just in case, you need to be certain. If you aren't... what would users think of your service?

Based on this, you can use an unambiguous icon (+ label, always use a label!) for cases where you're 100% sure there's an excess of usage. As we're at it, excess of usage is NOT an error, it's just a measure of usage, in any case you should let the user know they'll be charged for that, but it's nowhere close to an error.

On the other hand, if you're unsure what's going on, display some kind of warning explaining what's going on. In this particular case, you can be a bit ambiguous in order not to look as if you don't know what's happening. You could also mention possible causes, just like Google Chrome does:

enter image description here

Of course, this is not your specific case, but you'll notice the wording. Also, while not visible in the image, these messages have the technical information hidden by default, so consider that as well

1

The whole purpose of a dashboard is its ability to tell you how things are going, and where appropriate alert you to something that needs your intervention.

For dashboards Red is one of those colours (at least in western society) which screams "look at me" and as such it must be used sparingly, otherwise it risks getting lost in all the noise. It should be reserved for critical things, and because errors can vary in severity, e.g. some are minor and others are critical, so maybe you need an alternate grading system where only critical things which require your immediate attention are coloured red.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.