Great output, but IMHO the data source should change to be just INFO and not MONITOR: with Redis 2.4 you can't show top-commands. With 2.6 INFO has enough data about this (see INFO commandstats).
About top-keys, no way to show this without MONITOR, but it's not the most important bit of the visualization IMHO.
So very cool but a tool like this should not try to capture the stream of the commands executed by instead rely on what the system can provide as already aggregated stats.
agreed on MONITOR command, i rather just use INFO however on our deployment, top keys and commands are usually the most important information i'm looking for and with 2.4 there is currently no way.
When full monitoring is needed (like for the top keys) but still the analysis must be run continuously on a production instance, maybe in the next version of MONITOR (already planned) we could say to it "give me N items and return back in normal mode", or a similar thing but with time (for instance 1 millisecond), so that it is possible to perform some kind of polling that has no serious impacts on the server. Well I'll think about this when I'll reimplement monitoring :)
In redislive, the monitoring duration is configurable via command line argument, the idea is to schedule it as a cron job and monitor every 6 hrs for 1-2 minutes, that's enough to build the trend for analytics.
Worth browsing the code just to see some very well organized source for a tornado+redis+backbone stack, which is pretty popular I think on the real-time web.
I noticed you're using tornado but appear to be using the standard (non-evented?) python redis library. Does using this library block tornado's event loop? Did you look at other tools like brukva[1]?
The text for the upper boxes is easy to read on the full sized png, but difficult when scaled when showing in the github page. If the smaller size is the default it could be an issue.
I also think you should make scales for "Used Memory" and "Memory Consumption" the same.
sorry for not just digging through the source, but what are you using for the charts?
Are you storing metric data? or is it all in the browser? (e.g. do you get the full screen you just showed when you first open it up, or is it like a process mon where it's all 0 until you look at it)?
About top-keys, no way to show this without MONITOR, but it's not the most important bit of the visualization IMHO.
So very cool but a tool like this should not try to capture the stream of the commands executed by instead rely on what the system can provide as already aggregated stats.