Oracle Communications Performance Intelligence Center (PIC) Software
 

PLA-Support>Sun Systems>CommsGBU>Global Signaling Solutions>SN-SND: Tekelec PIC
 

In this Document

Symptoms

Changes

Cause

Solution

References

Created from <SR 3-9173098581>

Applies to:

Symptoms

Correlation rates of some reconstitution sessions dropped down. The number of "Not Matched" and "Timer" transactions (xDRs) considerably increased.

Changes

Cause

Manual reconstitution using the builder correlation keys showed synchronization issues between the different messages(PDUs) of a same transaction (xDR): Messages are not timestamped in the good order.

For example, when dealing with MAP protocol, END message is timestamped prior to the BEGIN message of the same transaction (xDR). Consequently, the builder wouldn't be able to correlate both messages: The BEGIN is marked as "Timer" while the END message is considered as "Not Matched".

This behavior is caused by NTP synchronization issues between the reconstitution chain servers (xMFs and IXPs).

The "ntpq -np" command shows the synchronization status between the concerned sever and the NTP source server. The most important parameters to check in the "ntpq -np" command result are the Delay (should be below 10ms) and the Offset (should be lower than 1ms).

[root@ixp1111-1a ~]# ntpq -np
     remote           refid      st t when poll reach   delay   offset  jitter
==============================================================================
*10.111.12.2    192.168.11.147    3 u  188  68m  377    0.445   -0.307   2.487
 127.127.1.0     .LOCL.          13 l   29   64  377    0.000    0.000   0.001

Note that the servers belonging to the same reconstitution chain must be synchronized to the same NTP source.

Solution

Synchronize all servers to the same NTP source. Use the following commands on the impacted servers:

1. Stop ntpd service. As root:
   [root@ixp0001-1a ~]# service ntpd stop
   Shutting down ntpd:                                        [  OK  ]
2. Force synchronization to the ntpserver1. As root:
   [root@ixp0001-1a ~]# ntpdate ntpserver1
   24 Jun 18:59:59 ntpdate[30298]: adjust time server 10.31.1.208 offset 0.005811 sec
3. Restart ntpd service. As root:
   [root@ixp0001-1a ~]# service ntpd start
   Starting ntpd:                                             [  OK  ]
4. Verify ntpq -np output. Start ("*") must be present after few minutes.

Where "ntpserver1" must be the same reference for all servers.

If one server is synchronised to the good NTP source but having some delay, you need to wait until synchronisation is done. This can take several hours to be done especially when the delay between host and peer is very big. In some cases, you might manually change the server time to adjust it as much as you can to the NTP source time (Once done, don't forget to restart the ntpd service). This should speed up the synchronisation.

References

This solution has no attachment