At work, as I’ve grown the metrics our production environment generates into the thousands, I’ve always wished for something like this that could (semi-) intelligently give a top-10 or something of “metrics acting weird.”
I guess when my kids are sysadmins some day and something like this is built into Nagios 24 I can reminisce about the good old days of digging through thousands of metrics by hand.
At work, as I’ve grown the metrics our production environment generates into the thousands, I’ve always wished for something like this that could (semi-) intelligently give a top-10 or something of “metrics acting weird.”
I guess when my kids are sysadmins some day and something like this is built into Nagios 24 I can reminisce about the good old days of digging through thousands of metrics by hand.
Also I love the interactive overview of DBSCAN clustering they reference.