TODO list
=========

- Add the possibility to list not only a unique monthly report for each LAN
  device defined, but a complete listing.

- Add the possibility to select between TCP and UDP when monitorizing network
  ports activity.

- Think an idea of how to solve the problem with All Host/All Graphs. Currently
  if you select such combination the All Graphs becomes as CPU load when using
  the Multihost feature. Yeah, I know, Mr. JavaScript.

- Add a daily counting for the CUPS jobs in the Network Services graph.

- Add the ability to change the size of the graphs.

- Add the ability to change the order of the graphs. For example, first the network graphs and then the CPU graph.

- Add the ability to toggle results in network graphs between MBs and Mbs (MegaBytes/s and MegaBits/s).

- Support for Dovecot pop3 logs in /var/log/imap.

- Someone reported that it could be interesting to monitor the NTP (ntpq -p).

- It could be useful to have a way to monitor a remote server without having
  installed Apache there. One possibility would be the ability to include paths
  in the %SERVER_LIST, in order to reach those names on a mounted remote fs.

- It would be interesting to monitor the I/O of other disks. The main problem is
  the layout.

- Notes about Smartmontools implementation:
	smartctl -A /dev/hda | egrep Temperature_Celsius | cut -c86-
	smartctl -A /dev/hda | egrep Temperature_Celsius | awk '{print $10}'
	and with "Current Temperature" !!!

- Add the ability to graph GPU temperature from nvidia-settings. This will
  require a redessign of the core temperatures graph. (by Audio Phile).

- Add the ability to not show a whole group of network ports if they aren't
  defined.

- CPanel FAQ, it could be interesting to add in httpd.conf the following lines:
  AliasMatch ^/?monitorix/?$ /var/www/html/monitorix
  ScriptAliasMatch ^/?monitorix.cgi/?$ /var/www/cgi-bin/monitorix.cgi

- Convert the @MNT_LIST into a %MNT_LIST in order to give the oportunity to
  rename the mount point either to have a shorter name or to hide its meaning.

- Add support for Xen hypervisor in the '/proc/interrupts' output:
------------------------------------------------------------------
           CPU0       CPU1       
256:  340379473          0  Dynamic-irq-level     timer0
257:     364739          0  Dynamic-irq-level     resched0
258:         36          0  Dynamic-irq-level     callfunc0
259:        368          0  Dynamic-irq-level     xenbus
260:          0     520182  Dynamic-irq-level     resched1
261:          0         62  Dynamic-irq-level     callfunc1
262:          0  340111345  Dynamic-irq-level     timer1
263:        160          0  Dynamic-irq-level     xencons
264:    1964186          0  Dynamic-irq-level     blkif
265:     682828          0  Dynamic-irq-level     blkif
266:    4175946          0  Dynamic-irq-level     eth0
NMI:          0          0   Non-maskable interrupts
RES:     364739     520182   Rescheduling interrupts
CAL:         36         62   function call interrupts
------------------------------------------------------------------
  Due that there timer interrupt has number 256 instead of 0, this will require
  to rewrite all the interrupt management in order to select them by name and
  then putting them into an array for each graph.

