NetXMS Support Forum

Hi!

Business services is a first step towards implementing SLA monitoring. Main idea is to give possibility to define services with criteria to determine availability of those services. Each business service can contain subordinate business services and/or checks used to determine service status. Business service can have only two statuses - up or down. If at least one subordinate service is down or at least one check reporting failure service is considered down. Availability for current day, week, and month calculated automatically and displayed in console. Also, there are ready to use report for service availability for any given period.
In it's current form configuring business services is very time consuming task. I'm thinking of possible ways to simplify this task. Also, in our plans is to introduce new object type - SLA (Service Level Agreement) to allow automatic SLA violation checking.

Best regards,
Victor

Hi!

Current situation with UNICODE is following:

console and communication protocol uses UNICODE, so no problems there;
agent does not use UNICODE;
server uses UNICODE only on Windows.

So most likely your problem is with server code page or agent code page. You can control server's code page by setting parameter CodePage in netxmsd.conf.

Best regards,
Victor

Hi all!

NetXMS version 1.1.10 is out. Changes from previous release:

- Completely new web interface
- New action type: execute NXSL script
- Changed default timeout for service checking subagent
- Added Oracle monitoring subagent
- Added option to force creation of character string instead of hex string in SNMP trap mapping
- Java console:
   - Log viewers greatly improved
   - Added missing DCI option "process all thresholds"
   - Added missing "interconnect networks" property for cluster objects
   - Added possibility to add and edit names on map link
   - SNMP trap monitor
   - "Query" button implemented in agent parameter selection dialog
   - Implemented syslog parser configuration
   - Resolved issue: Cannot delete non-existing object from map
   - Resolved issue: Template filtering script is lost when template is renamed
   - Resolved issue: Tab character not stripped from DCI parameter configuration and server reports "unsupported"
   - Resolved issue: Map background not drawn beyound viewport
   - Resolved issue: Node under cluster not shown in object selection dialog
   - Resolved issue: "Cluster resource" field unavailable for DCI on cluster node
   - Resolved issue: Reports not working on Windows
   - Resolved issue: extra new line characters in local command output
- New format specifiers in nxalarm: %x and %X
- New MIBs added: BAY-STACK-ADAC-MIB, BAY-STACK-ARP-INSPECTION-MIB,
  BAY-STACK-DHCP-SNOOPING-MIB, BAY-STACK-ECMP-MIB, BAY-STACK-ERROR-MESSAGE-MIB,
  BAY-STACK-LACP-EXT-MIB, BAY-STACK-MULTICAST-FLOODING-MIB,
  BAY-STACK-OSPF-EXT-MIB, BAY-STACK-PETH-EXT-MIB, BAY-STACK-RADIUS-MIB,
  BAY-STACK-SOURCE-GUARD-MIB, BAY-STACK-STATS-MIB, BAY-STACK-VRRP-EXT-MIB,
  FOUNDRY-CAR-MIB, FOUNDRY-SN-AGENT-MIB, FOUNDRY-SN-IP-MIB,
  FOUNDRY-SN-OSPF-GROUP-MIB, FOUNDRY-SN-ROOT-MIB,
  FOUNDRY-SN-SW-L4-SWITCH-GROUP-MIB, FOUNDRY-SN-SWITCH-GROUP-MIB,
  FOUNDRY-SN-TRAP-MIB, FOUNDRY-SN-VSRP-MIB
- Resolved issue: Subnet objects not placed correctly under zones
- Resolved issue: No retention time for alarm history

Best regards,
Victor

Hi!

ports used for monitoring are following:

udp 161 - SNMP requests - server sends request to port 161 on switch
udp 162 - SNMP traps - switch sends traps to port 162 on server

tcp 4700 - NetXMS agent - server connects  to remote host on port 4700
tcp 4701 - NetXMS management - management console connects to server port 4701

For SNMP devices you should use port 161 for requests going to device and 162 for traps going to server.

Best regards,
Victor

It is bundled with NetXMS installer. You don't have to install anything for using SQLite, just select SQLite as database engine in server configuration wizard.

Best regards,
Victor

Windows agent ne podderzivaet etot parametr. Ja dobavil eto v bugtracker, no ja poka ne znaju, kak poluchit' I/O wait v Windows i mozno li eto sdelat' v principe. Budu smotret'.

Hi!

You can use embedded SQLite database, then you will not need any external database engine.

Best regards,
Victor

Quote from: moseszero on February 06, 2012, 10:46:51 AM
HI,
I'm trying to create a DCI for a Node with Java API(version 1.1.6),and noticed that when create a DCI in Java console,after some DCI's parameter selected, its default information like data type,origin are also specialized dynamically,
how to get these default setting information with Java API or any other advice for creating DCI?

When you use DataCollectionConfiguration.createItem call, new DCI is created with the following default parameters:

origin - agent
data type - integer
polling interval - 60 seconds
retention time - 30 days
status - ACTIVE
delta calculation - none

all other attributes are set to 0, null, or empty string, depending on attribute's type.

You should set all required attributes and call DataCollectionConfiguration.modifyItem.

Best regards,
Victor

Hi!

Quote from: moseszero on February 07, 2012, 04:31:14 AM
Is the list of all supported DCI to some Node  available using Java API?

Yes, you can use Session.getSupportedParameters call. It accepts node object ID and returns list of parameters supported by that node.

Best regards,
Victor

Send it to [email protected]

Best regards,
Victor

Hi!

Could you please run netxmsd with option -D 9 and post the output?

Best regards,
Victor

Hi!

You mean high availability for NetXMS server itslef? You can check this topic: https://www.netxms.org/forum/configuration/netxms-in-a-cluster-mode/ for one possible approach to run NetXMS server in active-standby cluster.

Best regards,
Victor

Hi!

Yes, there are no command line tools for accessing SNMP devices via proxy. This functionality is intended for server to remote device communications. I never though that somebody will want to do this from command line It is possible to modify nxsnmpget/nxsnmpwalk tools to work via proxy agent. I'll add it to the list of feature requests, but I cannot promise exact version when it will be implemented.

Best regards,
Victor

Hi!

Yes, it's actually a bug. There should be retention time for alarms. I'll add it to bug list.
Currently, you can safely delete terminated alarms while NetXMS server is running. You can do it like this:

DELETE FROM alarms WHERE alarm_state=2 AND last_change_time<(timestamp here)

We use UNIX timestamp (seconds since 01.01.1970 00:00:00 UTC).

Best regards,
Victor

Hi all!

Attached are SNMP trap mappings for Dell and HP servers and APC devices, as well as data collection templates for Dell and HP servers.

Best regards,
Victor