<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"> <html> <head> <meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type"> </head> <body bgcolor="#ffffff" text="#000000"> Mohammed Morsi wrote: <blockquote cite="mid:4847EEE5.2030307@redhat.com" type="cite"> <meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type"> Mohammed Morsi wrote: <blockquote cite="mid:4847ED58.8010903@redhat.com" type="cite"> <meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type"> <title></title> Mohammed Morsi wrote: <blockquote cite="mid:48472729.2080005@redhat.com" type="cite">Attached in Mark's Monitoring API added to the app/util directory and my code interfacing it to generate the graphs. As far as I can tell everything works correctly, but due to my setup, I am getting alot of null datapoints back (this was verified though with the command line rrdtool). If all is well the graphs on the hardware summary page and hosts pages should be done. The graphs on the vm pools summary and vms page as well as the hardware vm pools page (a mirror of the summary snapshot stats for the selected pool) are not done as I am unclear as how to go about this, namely what statistics we need to collect for the vm pool and how to go about doing it (the monitoring API requires a host name, which is easy with hardware pools which contain hosts, but does not make much sense when it comes to vm pools). I can continue hacking away on this once this is resolved. -Mo <pre wrap=""><pre wrap=""><hr size="4" width="90%"></pre></pre> </blockquote> </blockquote> </blockquote> Attached is the updated patch which includes all of the previous plus the following: * graph styling / layout improvements and fixes * performance improvements, a few of the graphs on the summary page are now being preloaded and returned w/ the request (js still used but not to call back to the server) * VM Pool Graphs, temporary solution using the hosts associated with with vms in pool to collect stats implemented If everything looks good, essentially the only thing remaining with pre-summit graphing is to hook up the backend for the flexigrid load graphs (and more performance improvements if required) -Mo </body> </html>