I really like those ideas. The second one sounds dead simple to me, but it would actually be very useful. The first one I'm just trying to understand - are you saying you would define a check, and then easily assign multiple devices to it (rather than define a device and then assign checks to it?). If that's so it would be great!
I'm sure Dirk could whip those up in 15 minutes or so... ;-)
Ian
Ian K Gray
OEL IS - European Infrastructure Support
Tel: +44 1236 502661
Mob: +44 7881 518854
| "Vogl, Tom" <[EMAIL PROTECTED]>
Sent by: [EMAIL PROTECTED] 01/09/2005 23:16
|
|
Dirk - based on this - here's an idea for a new feature in SA:
Allow user to create a new check and use "[*]" as the "Server name or IPX address".
Then, upon SAVE, SA would pop up a list of all the "Server name or IPX addresses" it has already in the file and allow you to check one or more. It would then write out the new check for each server selected.
Ideally I could use the [*] string anywhere in the check definition and this process would replace it with actual name(s) selected. (ie: in the CHECK and ALERT tabs)
Obviously the "depends on" field couldn't be filled in for checks created this way, but I think most other fields could......
SECOND (AND COMPLETELY UNRELATED) IDEA:
A new status for checks: DISABLED.
Here is what I am thinking: DISABLED acts like MAINTENANCE in that checks are skipped each cycle.
BUT - DISABLED checks are also skipped (not processed) by all the output processes (web pages, interchange files, etc.). Basically you can't see them anywhere except in the GUI.
Let me explain my reasoning for this request with an example:
I have 4 fax servers in production (1,2,3,4). I have several checks for each that I have created. If one unit fails, I bring in a spare - it becomes unit 5 and I create checks for that unit in SA and unit 1 now becomes my spare offline unit (after it is fixed).
Rather than put all the old Unit 1 checks on MAINTENANCE and see them appear all the time - I want to hide them.
I have 4 units in production (2,3,4,5) and SA checks them and everything is fine. I want SA to completely ignore checks for unit 1.
I don't want to delete them from SA because I may want to bring unit 1 back on-line and I hate recreating all the checks.
The way SA currently works - I see a whole group of checks in "MAINTENANCE" for unit 1 - and it looks like I am dealing with ongoing operational issues. The actual fact is that all production systems are fine and there are no ongoing issues. The DISABLED checks for unit 1 can be quickly turned back on when I roll-out the unit again - but they disappear until then.
Think: Executives walk by and see all a completely "green" status board instead of one with lots of un-necessary yellow.
Does that make sense?
=tv=
Tom
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Troy Bruder
Sent: Thursday, September 01, 2005 3:58 PM
To: salive@woodstone.nu
Subject: [SA-list] SA "management"
Help,
How do others manage servers alive? Specifically, I have it setup that I perform a PING test of a server/device, and then if that passes, anywhere from 30-40 subsequent checks against that one server...
We have 15 servers in one of my environments.... and I've recently come up with 5 new monitors that I need to configure for each server.. however, it's painful to have to CTRL-D an entry, then change the name, pretty name, and depends on category for each of the 5 new monitors times 15 servers..
Is there an easier way??
Troy
____________________________________________
Troy Bruder, Manager – Hosting and Microsoft
Services
APT Information Technology Consultants
7540 Windsor Drive; STE 204
Allentown, PA 18195
Phone: 610-366-1100 –or- 888-2-GET-APT
www.aptconsulting.com
