1)Window resize. When new service added the window gets wide again and must be
resized back down. (EricH)
From BradG: priority From KenM: Can we be more specific? You start with a window 2 inches wide, to just fit the selected services. It seems reasonable to, if you add a service, enlarge the window to fir the new service?? Is the desired action to set a flag if you resize the window, preventing any further automatic 2 inches wide, to just fit the selected services. It seems reasonable to, if you add a service, enlarge the window to fir the new service?? Is the desired action to set a flag if you resize the window, preventing any further automatic window size adjustments? From JustinP: i think the main issue is the windows don't retain there size which is probably a result of no scroll bars. things are setup to be automatic currently but i'm guessing the consensus is to change that to panes which use scroll bars and leave the window alone, set at whatever the user has made it. this can be saved in the prefs file to persist across restarts. if a ton of new things are added they just go off the bottom/side (pick one) and the scroll bar is there to let you know that happened and indicate there's more to see if you care to resize or scroll the window. From KenL:
I feel their pain. I had Fox add an Xresource to allow one
to specify a maximum width. Addition (and removal) of services
won't exceed that width. By default it uses the screen width
but I've got a little launcher window and a new mail indicator
that I keep in the upper left that I don't want Argus occluding. | ||||||||||||
2) Web interface for places where full client not available. (EricH) | ||||||||||||
3)
Popups steal focus so they interrupt typing in another window. (EricH)
From KenM: Is this a bug or feature? I agree it can be annoying at times, but when I have switched to another virtual screen and something pops up on the window window where argus is running, I might want to know about that. I think it's essential for Ops's dashboard to work like that. Perhaps a '-grabfocus' option? From Brad: priority From Ken: Fvwm allows me to specify which windows require "click-to-focus". I've set all the Argus popups to require a click. It's most disconcerting to have an Argus popup show up while you're in the middle of a sentence only to have the rest of that sentence discarded into the bitbucket. Then you've got to reach off the keyboard to find the mouse to get your window back and by then you've lost your thought. It would be really irritating if the popped up window isn't even anywhere near the cursor. Argus doesn't need focus (you can't type into it) so stealing the focus seems obsessive. I don't even think it would be good for ops. Pop up on top, maybe, but steal the focus, no, not even if it's under the mouse pointer. From RickW: I'll vote to default to modal grabs and add a --nograbfocus option. | ||||||||||||
4)If main window is at the bottom edge of the screen then all popup windows
pile up on top of each other at the bottom or off the bottom of the screen. I
suppose this is because they always try pop up below the main window on the
screen. (EricH)
From Brad: priority | ||||||||||||
5)A way to only show services/hosts which my group is responsible for. A way to see all services/hosts some other group is responsible for. (EricH) | ||||||||||||
6)A way to only show specific hosts which I care about. (EricH) | ||||||||||||
7)Each person shouldn't have to set their own list of services to watch but be abe able to select a common set. (EricH) | ||||||||||||
8)Small view which only shows exceptions (non-green.) (EricH) Alternate argus view that just show a list/table of exceptions (BradG) From KenM: In progress From Brad: priority | ||||||||||||
9)A view which shows the problem description without having to pop up a little window. (EricH) | ||||||||||||
10)A mode which doesn't pop up any little exception windows. (EricH) | ||||||||||||
11)View excetions by host or color. (EricH) | ||||||||||||
12)Summary view of a host which shows all services being monitored on that host and their status, regardless of color. (EricH) | ||||||||||||
13)Doesn't work right when off campus for for some class of machines. Maybe
UW-only addressed. I haven't tested this in a while. (EricH)
From KenM: Can you retest (works OK for me at home, FWIW). | ||||||||||||
14)Different users may have different information depending on when they read
the rc file. Eliminate the need to restart client to get the most up to date
setup. (EricH) reduce need for argus restarts (BradG) From Brad: priority
15)When new services are added there is no mechanism for users to learn about
them. Perhaps make new services appear by default and must be turned off. (EricH)
|
16)Be able to setargus a service/host from within the Argus client. (EricH)
|
17)A setargus state beyond just up/down or color change action which will
indicate that a problem has been acknowledged so it need not remain red while
the problem is being resolved, nor require it to be marked "down" to squelch
the red. (EricH)
|
18)When you want to change multiple fields in ARS,
and you're setting Host Monitoring to or from yes, there is a restriction on
how you do the ARS change. I believe that that is for interoperability with
the argus clients and their reload of the xhm stuff.
I'd like to get rid of the restriction on how you do the ARS change. (BobS)
|
19)With the default Microsoft Windows "Internet Options/Security" settings,
running C:\program files\argus\argus.bat causes a security warning.
"The publisher could not be verified...". I'd like to have argus start
without the warning, and without otherwise loosening Internet
Options/Security. Even if it is the first time I'm running java argus on a
PC. (BobS)
|
20)Prioritize pop-ups (BradG)
| From KenM: to elaborate (I think): this means the ability to have alerts for certain services prioritized in some manner (top of the pile, or just suppressing all others for a host). The idea is to have, say, a box being totally down be reported as 'XHM for cg66', not 'CLOCKD for cg66'.
21)Filter events based on time (BradG)
| From KenM: to elaborate (I think): this means the ability to have certain alerts (by host:service:severity) suppressed during certain times.
22)2 color view for ops (hide all blue, green, brown items) (BradG)
| From KenM: In progress From Brad: priority
23)merge info/event with pilot viewer (so 1 viewer tool instead of 2 (BradG)
|
24)Not have java argus pop-up window that are partially or fully off screen or
covering the main window (SteveW)
|
25)Present an error message instead of silently failing if java runtime
environment is damaged or missing. Probably an installer package on windows
that would gripe if your java installation isn't working. (SteveW)
|
26)Client version information visible from the client (SteveW)
| |
1)Proxy setup is a mess and not very scalable. Can a more elegant solution be
found?
From KenM: What he said!!!!! I dunno if that's a near term thing, though. (EricH) |
1)Systems self register, no server side config (rc) required.
found?
(EricH)
From KenM: Bug or feature?
|
2)Status changes made in one client are visible to all clients. Notably the
change of grade to acknowledge an event. This makes it more collaborative
(EricH)
From KenM: I may not be following. The argus status changes (red->green, green->red, red->white (when setargus'd)) are reflected to all clients. If we added mauve for acknowledged, that would be reflected also. |
3)Report informational stuff like serial number, hardware type.
(EricH)
From KenM: for a partial solution, the proposed 'system health' pilot client can report that in a 'full' report, or somesuch. |