1.29. Release Notes 1.0.2¶
1.29.1. New features¶
- This version introduces discovery and monitoring of MPLS LSPs statistics. Currently we discover LSP on Juniper devices and monitor traffic (in bits/sec), packet rates and operational state of the LSP.
1.29.2. Improvements in the Server¶
This version comes with improvements in memory utilization and speeds up monitoring data processing in the Python hook script. It takes between 8 and 9 sec to process 350,000 variables on the system with >1000 devices
New function mvar_abs() has been added to the Python module nw2functions. This function calculates absolute value of the last observation in the time series of the input monitoring variable and puts it back into the same monitoring variable. Typical use:export_var('FooRate', mvar_abs(derivative(import_var('Foo'))))
this version introduced self-monitoring of the SQL database connections pool. Monitoring variables appear in the Graphing Workbench under category Monitor and are as follows:
- c3p0NumBusyConnections: Number of busy connections in the database connection pool
- c3p0NumConnections: Total number of connections in the database connection pool
- c3p0NumIdleConnections: Number of idle connections in the database connection pool
- c3p0FailedCheckoutsRate: Rate of connection checkout failures in the database connection pool
- c3p0FailedIdleTestsRate: Rate of idle connection test failures in the database connection pool
non-zero value of c3p0FailedCheckoutsRate or c3p0FailedIdleTestsRate indicates the pool manager has failed to open (“check out”) new database connection. When this happens, value of c3p0NumConnections becomes zero and stays at this level until connection to the database is reestablished.
1.29.3. Changes in the Monitor¶
Monitor will always send all batch queries to the device even if it takes longer than configured polling interval. If this happens, monitor makes a warning log record to report it.
implemented ability to set class used to build bulk SNMP GET queries via config file parameter monitor.snmpBulkQueryBatcher. Three options are available:
- BulkGetQueriesBatcher all OIDs are grouped equally
- BulkGetQueriesBatcherDot3AdOperStateAlwaysFirst one dot3adAggPortActorOperState OID is always placed first in each bulk query
- BulkGetQueriesBatcherDot3AdOperStateAlone each dot3adAggPortActorOperState OID is sent in a separate SNMP GET query
Default value is BulkGetQueriesBatcherDot3AdOperStateAlone