Tool to monitor HM / RMA

Need new test, action, option? Post request here.
Robert_in_MTL
Posts: 229
Joined: Tue Jun 20, 2006 1:20 pm
Location: Montreal, Quebec

Post by Robert_in_MTL »

Don't read me wrong! Having a second instance of HostMonitor is a working solution!

But I think it is awkward to install and manage another HM with another Serial number for this.
On the other hand, this solution works and you don't have to program anything to make this work.

I really think that it would be a lot simpler to everyone to manage only one instance. Only one place to login and manage everything, no fuss.

But it's your company and your software, you decide what's best for you and I' will take your solution.

But please take a step back and look at the whole picture:
1- having one instance doing it all.
2- having 2 instances to monitor each other.

From a marketing / selling point, it would appear as a flaw I guess, But then maybe I am wrong.

I hope you will reconsider.

have a great day Alex :D
User avatar
greyhat64
Posts: 246
Joined: Fri Mar 14, 2008 9:10 am
Location: USA

Post by greyhat64 »

Robert_in_MTL, I like the way you think! :D

Alex,
First of all you know I'm completely sold here - a great product with great support for an unbelievable price! I know, as well, that the focus lately has rightfully been on the core product (v8 release soon!). BUT. . .

If you are intent on selling to the enterprise, a primary goal must be to ensure the reliability of your own product while ensuring the reliability of the devices/services you monitor.
Of course you can provide a 'lite' license to allow administrators to do this for themselves, but we don't need 60 test types for AHM. Additionally, this approach lacks the elegance that a more focused solution would have in order to make it a low-maintenance, better integrated, self-correcting system. To do that you would need (in my opinion):
  • An outbound channel of communication directed at the watchdog, in much the way that Active RMA agents communicate to RMA Manager, would provide similar 'Not connected' or 'Operable' statuses for the Host server.
    -The added benefit to port level comms between the two is that legitimate changes to the Host system - ones that might affect monitoring the host (IP Address, administrator contact info, log file type, etc.) - could be communicated to the watchdog instead of depending on the administrator to make these changes to the 'lite license' config manually.
    - This would also allow the info to be 'pushed' across a firewall to the 'watchdog' if necessary.
  • Other tests, such as Service, ping, URL (for web services if enabled), and log file (or log query for DB logs), could be triggered to provide additional diagnostic information.
    - OK, so there are a number of tests listed here :oops:, still - it's not 60.
I hope you will at least consider this for a (near) future enhancement of v8. It will give AHM an additional appearance of maturity that might attract more attention ($$$) to the product. (Feedback from peers, collegues and clients I've pointed your direction)

If you are still intent on the 'lite license' approach, at least include instructions for the uninitiated (in the Help would be fine) or, better yet, a clear and discernable template for monitoring AHM. 8).

Sidenote: As for myself, I'll soon be setting up a 2nd Enterprise license (Two divisions that can't 'meet in the middle' :x ), so for now I'll have one monitor the other. Given the choice though, I'd still prefer an independant watchdog.
:roll: Which brings up another, maybe obvious, watchdog feature - the ability to monitor multiple Host servers if necessary.
KS-Soft
Posts: 13012
Joined: Wed Apr 03, 2002 6:00 pm
Location: USA
Contact:

Post by KS-Soft »

I really think that it would be a lot simpler to everyone to manage only one instance. Only one place to login and manage everything, no fuss.
Not sure I understand. How you can use 1 instance to monitor itself??
In any case you should install some software on 2nd system, in any case you should setup some alerts for that software, in any case you should tell how to connect to 1st instance.
The only difference name of that software: HostMonitor or Watchdog.
Well, we could add some options to simplify configuration of the utility but it will save you 5 minutes (not 5 minutes everyday, just 5 minutes).
That's why we still have this task in the list but this is low priority task.

Regards
Alex
KS-Soft
Posts: 13012
Joined: Wed Apr 03, 2002 6:00 pm
Location: USA
Contact:

Post by KS-Soft »

PS

1) If some option is not implemented but there is easy and quick workaround, such tasks always have low priority.
I think you spent more time for this discussion than you need to setup 2nd instance of HostMonitor to check 1st one :wink:

2) That's why "clustered monitoring" task has higher priority - there is no easy workaround.

Regards
Alex
Robert_in_MTL
Posts: 229
Joined: Tue Jun 20, 2006 1:20 pm
Location: Montreal, Quebec

Post by Robert_in_MTL »

Hi Alex,
Sorry, I was continuing from my previous post...
could we just select 1 or 2 remote monitoring agents so that they would check on HostMonitor on a regular schedule (every 30 seconds or every minute, whatever) and if they cannot reach HM, they would send admins an alert email.

I woudn't mind configuring a RMA for an SMTP server and a mail Template (but the best would be to push those settings from RMA Manager)

You would need to enable a new version of RMA to monitor HM on it's own and enable it to send a pre defined mail on it's own. That's it!

We would then just have to go to options and select which agent(s) is(are) designated to be so-called watchdog(s), HMwould then watch them back.
When I said
I really think that it would be a lot simpler to everyone to manage only one instance. Only one place to login and manage everything, no fuss.
I was referring to my previous post... Sorry if I was unclear.

I meant to log in HM once then choose which RMA will be the watchdog(s) to watch HM back. That's what I meant by login only in one place.

Hope it clarifies my suggestion.
KS-Soft
Posts: 13012
Joined: Wed Apr 03, 2002 6:00 pm
Location: USA
Contact:

Post by KS-Soft »

I meant to log in HM once then choose which RMA will be the watchdog(s) to watch HM back. That's what I meant by login only in one place.
Sorry, my fault :oops: Yes, it has sense for Active RMA. Actually we have similar task in "to do" list: "start action when agent cannot connect to HostMonitor within N minutes" option for Active RMA

Regards
Alex
Robert_in_MTL
Posts: 229
Joined: Tue Jun 20, 2006 1:20 pm
Location: Montreal, Quebec

Post by Robert_in_MTL »

Excellent!!!
That's what I had in mind.

Perfect! thank you Alex! :D :D :D :D
KS-Soft
Posts: 13012
Joined: Wed Apr 03, 2002 6:00 pm
Location: USA
Contact:

Post by KS-Soft »

greyhat64,

I agree with you. However we have 1000 tasks in the list (including this one) and you cannot convince me this is high priority task.
"If some option is not implemented but there is easy and quick workaround, such tasks always have low priority"

Regards
Alex
User avatar
greyhat64
Posts: 246
Joined: Fri Mar 14, 2008 9:10 am
Location: USA

Post by greyhat64 »

Robert_in_MTL, Once again I think you and I are on similar tracks.
It's not that it can't be done with a 'lite' license, and for some environments that may suffice. But more so, is that the best way to do it?

Alex, you may be right, we've spent a lot of time on this one, but look at how many hits this topic has (nearly 7,000!!!) I know it's a two year old thread, but I think there is a silent crowd that is interested in this, I just wish there were more participants in the conversation to convince you otherwise regarding this priority. (hint, hint everyone!).
Paul_NHS
Posts: 59
Joined: Wed Feb 25, 2009 6:17 am

Post by Paul_NHS »

My 2c.

I had in mind this scenario for monitoring HostMonitor.

Network: Multiple sites, multiple IP ranges. (Our current set up.)
Current Monitoring: 1 x HM sending email/page/SMS.
Proposed Monitoring: 1 x HM at 2 x sites. 2nd email server/pager/SMS modem at 2nd site.
  • HM at main site performs "normal" monitoring with backup SMTP server set to point to 2nd site email server.
  • HM at main site copies HML file to 2nd site regularly.
  • HM at 2nd site monitoring HM process/log file.
If HM at main site is down, HM at 2nd site performs optional additional test(s) to ensure main site is down, not 2nd site. Then runs HM script to load main test HML file at 2nd site.

Requires a second HM license, but that is cheap insurance.

cheers, Paul
Robert_in_MTL
Posts: 229
Joined: Tue Jun 20, 2006 1:20 pm
Location: Montreal, Quebec

Post by Robert_in_MTL »

Thank you for joining the topic Paul_NHS :D,

Good input, if this method is built in HM then it would be a great idea!:wink:

If not, I think that the drawback would be an increase in IT management and debugging.

Folowing your thoughts:
If in options, we could choose: ( [ ] Main Monitor [x] Watchdog/Backup ) and by choosing Watchdog/Backup we would only need to point to the Main HM, enabling the second one to do what you suggest. This would be Marvelous!

This is the only way that I would agree to install a second instance of HM.

For now, having the ability to choose which 1 or 2 or 3 RMAs will monitor back HM is the most appealing solution in my humble opinion.
KS-Soft
Posts: 13012
Joined: Wed Apr 03, 2002 6:00 pm
Location: USA
Contact:

Post by KS-Soft »

Paul_NHS,

Some people configured HostMonitor in such way. Its not very easy but it works.
On the other hand, I hope we implement clustering monitor in the future.

Regards
Alex
Paul_NHS
Posts: 59
Joined: Wed Feb 25, 2009 6:17 am

Post by Paul_NHS »

Yes, my suggestion is poor man's clustering and it is cumbersome, but better than not having a backup monitor.
I'm not sure that clustering is the best mechanism, as you should use an alternative site for monitoring - assuming you have multiple sites.

Monitoring should be relatively simple and self contained, to allow easy recovery/cut-over if required.

cheers, Paul
Scooby
Posts: 10
Joined: Wed May 21, 2008 6:46 am

Post by Scooby »

Hrmm...

My solution was to throw together a watchdog application that I had needed for in other areas anyway. HostMonitor checks in with the Watchdog application every 60 seconds. If the Watchdog hasn't heard from HM for 5 consecutive minutes, it starts sending out alerts. In turn, I have HM checking that the WatchDog daemon is running and processing jobs.

In the end, I have a system that watches after itself as well as the capacity of watching system availability as well as verifying scheduled events happen on time.

...Kinda fixes that question of who watches the watch men.
Paul_NHS
Posts: 59
Joined: Wed Feb 25, 2009 6:17 am

Post by Paul_NHS »

What did you throw the watchdog together with?

cheers, Paul
Post Reply