Koozali.org: home of the SME Server

Monitoring Multiple SME Boxes Remotely

Offline Gert

  • ****
  • 208
  • +0/-0
    • http://www.huge.co.za
Re: Monitoring Multiple SME Boxes Remotely
« Reply #30 on: June 23, 2008, 05:03:56 PM »
Yes please.

Offline mercyh

  • *
  • 824
  • +0/-0
    • http://mercyh.org
Re: Monitoring Multiple SME Boxes Remotely
« Reply #31 on: June 23, 2008, 05:05:44 PM »
I will work on it when I get time. I may PM you with the link so you can check for accuracy.

RH

Offline Gert

  • ****
  • 208
  • +0/-0
    • http://www.huge.co.za
Re: Monitoring Multiple SME Boxes Remotely
« Reply #32 on: June 23, 2008, 05:07:13 PM »
Ok, no problem.

Offline mercyh

  • *
  • 824
  • +0/-0
    • http://mercyh.org
Re: Monitoring Multiple SME Boxes Remotely
« Reply #33 on: June 23, 2008, 09:33:39 PM »

Offline sal1504

  • ***
  • 149
  • +0/-0
Re: Monitoring Multiple SME Boxes Remotely
« Reply #34 on: June 25, 2008, 05:10:18 PM »
Gert
Program looks excellent and I will be using it, but I have a small problem. On one of my servers I have a raid 5 with 4-250gb sata drives running SMEServer 7.3. I lost one of the hard drives yesterday, but NRSM did not detect or notify me of this lost drive. NRSM still shows the raid as clean. I did a refresh and it still shows the raid as clean. Any suggestions on how to fix.

Sal

Offline Gert

  • ****
  • 208
  • +0/-0
    • http://www.huge.co.za
Re: Monitoring Multiple SME Boxes Remotely
« Reply #35 on: June 25, 2008, 10:24:14 PM »
Sal

NRSM initially saves a snapshot of /proc/mdstat in its database. Mine looks like this:

Code: [Select]
md3 : active raid5 sdd1[3] sdc1[2] sdb1[1] sda1[0]
      2930279424 blocks level 5, 256k chunk, algorithm 2 [4/4] [UUUU]

md2 : active raid1 hda2[0]
      244091520 blocks [2/2] [UU]

md1 : active raid1 hda1[0]
      104320 blocks [2/2] [UU]

unused devices: <none>

With every refresh it compares the current /proc/mdstat with the saved one. Any change, even the slightest difference will raise a Failure warning. You can test this by opening the servers.mdb with M$ Access and make a small change in the RAID_Status field. NRSM will most definitely raise a Failure. I am confident that you lost your drive before you accepted the initial value of /proc/mdstat. An example of a lost drive would look something like this:

Code: [Select]
md3 : active raid5 sdd1[3] sdc1[2] sdb1[1] sda1[0]
      2930279424 blocks level 5, 256k chunk, algorithm 2 [4/4] [UUUU]

md2 : active raid1 hda2[0]
      244091520 blocks [2/1] [U_]

md1 : active raid1 hda1[0]
      104320 blocks [2/1] [U_]

unused devices: <none>
« Last Edit: June 25, 2008, 10:29:18 PM by Gert »

Offline sal1504

  • ***
  • 149
  • +0/-0
Re: Monitoring Multiple SME Boxes Remotely
« Reply #36 on: June 26, 2008, 12:00:43 AM »
Gert

Thanks for the quick reply. I apparently had the hard drive failure prior to installing NRSM and when it showed the raid configuration I selected OK. This is where I went wrong because I didn't look at the raid configuration. One other feature I would really like to see is NRSM to show the OS and OS Version. This would help me keep up with which machines need to be updated. I have 67 servers that I try to keep up with and this program will be of great help. Thanks for the time and effort you put into this

Sal

Offline kryptos

  • ****
  • 245
  • +0/-0
Re: Monitoring Multiple SME Boxes Remotely
« Reply #37 on: June 26, 2008, 05:14:16 AM »
Gert,

Does it only work if you have a raid setup? I tried other server with no raid it always says offline but with the server with raid it works fine.

Regards,
Rocel

Offline Gert

  • ****
  • 208
  • +0/-0
    • http://www.huge.co.za
Re: Monitoring Multiple SME Boxes Remotely
« Reply #38 on: June 26, 2008, 08:23:30 AM »
Does it only work if you have a raid setup? I tried other server with no raid it always says offline but with the server with raid it works fine.

Kryptos, you don't need a raid setup in order to monitor a server. Just make sure that remote ssh is enabled on the server and that it accepts standard passwords.

Offline Gert

  • ****
  • 208
  • +0/-0
    • http://www.huge.co.za
Re: Monitoring Multiple SME Boxes Remotely
« Reply #39 on: June 26, 2008, 08:28:07 AM »
One other feature I would really like to see is NRSM to show the OS and OS Version. This would help me keep up with which machines need to be updated.

Would there be more people interest in that feature? If so I would consider implementing it.

Offline kryptos

  • ****
  • 245
  • +0/-0
Re: Monitoring Multiple SME Boxes Remotely
« Reply #40 on: June 26, 2008, 11:59:07 AM »
Kryptos, you don't need a raid setup in order to monitor a server. Just make sure that remote ssh is enabled on the server and that it accepts standard passwords.

what you mean remote ssh enabled? I have enabled ssh but local connections only and the pc i use is in the same local network.

Regards,
Rocel

Offline Gert

  • ****
  • 208
  • +0/-0
    • http://www.huge.co.za
Re: Monitoring Multiple SME Boxes Remotely
« Reply #41 on: June 26, 2008, 12:37:30 PM »
Let me put it this way. You must be able to connect to the server from putty using a username and password. If the ssh is running on port 22 (the default) then you can leave the SSH Port field blank in Server Maintenance. Otherwise enter the post number. Please let me know if you come right.

Offline mercyh

  • *
  • 824
  • +0/-0
    • http://mercyh.org
Re: Monitoring Multiple SME Boxes Remotely
« Reply #42 on: June 26, 2008, 02:50:56 PM »
Gert,

If we are doing New Feature Requests here 8), I would like to see a date on the backup field. (Date and/or time of last successful backup.)

Offline sal1504

  • ***
  • 149
  • +0/-0
Re: Monitoring Multiple SME Boxes Remotely
« Reply #43 on: June 26, 2008, 04:58:13 PM »
Gert

I agree with MercyH. That is would be a great feature also.

Sal

Offline jdavey

  • **
  • 43
  • +0/-0
Re: Monitoring Multiple SME Boxes Remotely
« Reply #44 on: June 26, 2008, 07:11:03 PM »
Quote
Gert,

If we are doing New Feature Requests here , I would like to see a date on the backup field. (Date and/or time of last successful backup.)

Since you are doing all the heavy lifting - and now it appears, taking requests - in addition to the last successfull backup, about displaying the next scheduled backup?

This looks like it will become a full-time job....
...