IMonitor Diagnostic Logger


IMonitor Diagnostic Logger was a new Feature in Imonitor EDirectory (?????) and is still poorly documented and a sloppy implementation.

iMonitor click on the Agent Configuration - Diagnostic Logger links in the left side frame.

Diagnostic Logger Configuration

  • Logging Interval OK (secs) 120 - logging interval will normally be this in seconds
  • Logging Interval WARNING (secs) 10 - When a Lock recently used this on EDirectory (40004.44) and these are the Columns which are poorly laid out for Spreadsheet or analysing work. Here is our guide:
  • A - Current Time -
  • B - Status - "OK", "WARNING", "???"
  • C - Poll Count - increments the amount of the setting in Logging Interval each for each row
  • D - Logging Interval (sec) - Logging Interval as entered in iMonitor
  • E - Up Time (d:h:m:s) - Poorly updated value
  • F - Agent Load Time - When Agent was loaded (05/21/2019 12:41:21)
  • G - Agent Open Time - When agent opened (05/21/2019 12:41:21)
  • H - Agent State - "Open"
  • I - DIB Locks (Total) - Total DIB Locks within the Logging Interval
  • J - DIB Lock Holder (0xThread ID) - (0xDD40C700) Does not appear useful
  • K - DIB Lock Holder (NDS Thread ID) - (0xDD40C700) Does not appear useful
  • L - DIB Lock Holder (NDS Thread desc.) - "SkulkerWorkerProc", "DSAEndUpdateReplica", "DSAUpdateReplica", "DSAModifyEntry"
  • M - DIB Lock Holder (Hold time - sec)
  • N - Idle Pool workers
  • O - Total Pool Workers
  • P - Peak Pool Workers
  • Q - Current Ready Workd
  • R - Peak Ready Work
  • S - Current Waiting Work
  • T - Peak Waiting Work

Comments on Ldapwiki experiance#

  • Column A shows the DIB Lock began on the 15th of June around 3:36 pm and ended at 3:41 pm. More significantly it ended on its own since column E - Up Time (d:h:m:s) shows the server was not rebooted and column G - Agent Open Time, shows the daemon was never restarted.
(The time display is approximate and does not increment properly in this first version.)
  • Column J - DIB Lock Holder (0xThread ID) and K - DIB Lock Holder (Thread ID) show the thread's ID in hex and decimal. These are not terribly useful on Linux since there is no way to cross reference these to the threads found in a core if one is available taken.
  • Column L - DIB Lock Holder (Thread desc.) which shows that in this case the Janitor process was holding the lock and held it for 323 seconds as indicated by column M - DIB Lock Holder (Hold time - sec).

This log may be useful to Novell Support. If using this functionality for a company's own investigation the following questions might be asked: 1. Does the lock time always occur around the same time of day? 2. What is happening in the environment during this time? 3. What has changed in the environment? 4. Grab some dstrace output before the issue occurs to see what activity leads up to the issue.

It is important to be aware that currently after eDirectory is restarted or the server is rebooted that the diagnostic logger settings will revert to off. An enhancement request has been entered to make this permanent if desired.

Diagnostic Logger Configuration

  • Logging Interval OK (secs) 300
  • Logging Interval WARNING (secs) 30
  • Lock Threshold (secs) 180

As seen at iMonitor to troubleshoot eDirectory database locks When things turn bad, the "Status" column B changes from "OK" to "WARNING".

All of out entries show "OK".



More Information#

There might be more information for this subject on one of the following: