Mutiple met_id are used for same net_connect metric

Products

• DX Unified Infrastructure Management (Nimsoft / UIM)

Issue/Introduction

In data_engine log:

Aug 13 01:35:16:421 [25724] de: Metric id to be changed from M25223241FC173E957059DDCAE9802C61 to M8E3992C2CDDD141FDECE73211C3DB8F1 for qos=QOS_NET_CONNECT_PACKETLOSS source=xx target=yy:packet_loss robot=zzprobe=net_connect queue size= 0

 ...

 Aug 13 01:35:16:422 [25724] de: Metric id to be changed from M8E3992C2CDDD141FDECE73211C3DB8F1 to M71A8859A99BB7CE1CD1C7ADDD87EB623 for qos=QOS_NET_CONNECT_PACKETLOSS source=xx target=yy:packet_loss robot=zz probe=net_connect queue size= 1

 ...

 Aug 13 01:35:16:422 [25724] de: Metric id to be changed from M71A8859A99BB7CE1CD1C7ADDD87EB623 to M25223241FC173E957059DDCAE9802C61 for qos=QOS_NET_CONNECT_PACKETLOSS source=xx target=yy:packet_loss robot=zz probe=net_connect queue size= 2

...

Aug 13 01:35:16:704 [23824] de: [QOS Metric] has connected to database Aug 13 01:35:16:705 [23824] de: qos_metric_thread - updated qos object with table_id 42 with metric id 'M8E3992C2CDDD141FDECE73211C3DB8F1' time used open db: 13 ms, execute cmd: 1 ms Aug 13 01:35:16:707 [23824] de: qos_metric_thread - updated qos object with table_id 42 with metric id 'M71A8859A99BB7CE1CD1C7ADDD87EB623' time used open db: 0 ms, execute cmd: 1 ms Aug 13 01:35:16:708 [23824] de: qos_metric_thread - updated qos object with table_id 42 with metric id 'M25223241FC173E957059DDCAE9802C61' time used open db: 0 ms, execute cmd: 1 ms

Client application 'data_engine  [QOS Metric]' will update the s_qos_data for the same table_id:

update s_qos_data set ci_metric_id = 'M8E3992C2CDDD141FDECE73211C3DB8F1' where table_id = 42;

update s_qos_data set ci_metric_id = 'M71A8859A99BB7CE1CD1C7ADDD87EB623' where table_id = 42;

update s_qos_data set ci_metric_id = 'M25223241FC173E957059DDCAE9802C61' where table_id = 42;

Environment

data_engine probe 8.50 and later with parameter 'update_metric_id' set to 'yes'

net_connect 3.33 with packet loss metric enabled

Cause

A known net_connect issue that was fixed with net_connect 3.37

Resolution

deploy net_connect 3.37.

drop niscache folder and restart nimbus service on the robot server to which net_connect probe is deployed.

Check data_engine at log level 5, messages as in the description will not occur again. Profile on the DB table s_qos_data, the update statements should not be seen repeatedly for net_connect probe.