NetXMS Support Forum

I've made some changes in installer - hopefully it will help. Please try 2.0-M2 when it will be available.

Best regards,
Victor

Hi,

try to use CPPFLAGS instead of CFLAGS - CFLAGS only affect C sources, while CPPFLAGS affects both C and C++.

Best regards,
Victor

Hi,

I likely fixed it in development branch. Fix will be part of 2.0-M2 release (will be published very soon). It works on my CentOS 6.5 test system.

Best regards,
Victor

Yes, upgrade procedure is the same - you just run nxdbmgr as usual and it will make necessary changes in DB.

Best regards,
Victor

Yes, it's exactly same bug - NetXMS was not read ipAddressTable. After upgrade to 2.0-M1 or newer your problem should be fixed.

Best regards,
Victor

Hi!

1. In Last Values for cluster you will see only those DCIs that are marked for aggregation (see Cluster property page of DCI) - and for them you'll see aggregated value (for example, average COU usage for cluster nodes).

2. Yes, it's not implemented yet.

3. Yes, it will be published soon. It is in my short TODO list.

4. Agree.

5. Cluster DCI configuration works like a template that is applied to all cluster nodes. So any DCI without resource association will be collected on each node.

6. Currently it's not possible to create template with cluster resource binding. I'll think how it could be implemented.

7. Interfaces on cluster nodes with addresses belonging to cluster networks excluded from network topology. It is quite common setup when multiple clusters has identical IP addresses on interconnect links.

Best regards,
Victor

Quote from: VMGuy on January 13, 2015, 10:23:59 PM

QuoteCheck agent's config file – server IP address should be whitelisted by using "MasterServers" option, all others ("Servers" and "ControlServers") don't have sufficient rights to edit configuration file remotely.

In the config file I have (and always had) this:
Servers = 192.168.x.y
ControlServers = 192.168.x.y
MasterServers = 192.168.x.y

But it doesn't work.

That's new bug introduced in 2.0 - server's access level determined by first statement insted of one with maximum access. If you leave only MasterServers (which implies ControlServers which in turn implies Servers) it will work.

Best regards,
Victor

There was very similar problem on VMWare ESX which was fixed in 2.0-M1. What NetXMS version you are using?

Best regards,
Victor

no, custom attributes are not intended to be copied when you apply template. You can use scripts to create custom attributes where needed (for example, handle SYS_NODE_ADDED event, identify node as SNOM phone and add this attribute).

Best regards,
Victor

Hi,

I was able to reproduce problem with netxmsd.exe hang (on Windows Server 2003 32 bit). I'll publish installer with a fix later today.

Best regards,
Victor

Can you please change LogFile from {syslog} to some file and run netxmsd -D9 from command line? It should show at least some messages.

Best regards,
Victor

Hi,

can you please show your netxmsd.conf?

Best regards,
Victor

Yes, it's a bug in installer. I just uploaded installer with a fix, please try it (file names are the same).

Best regards,
Victor

Hi all!

NetXMS 2.0-M1 is released. This is first release in 2.0 branch. From now on we change version numbering scheme. Now it consists of only two numbers followed by qualifier. Qualifier could me M (for Milestone) which indicates milestone toward release, RC (for Release Candidate), and no qualifier for stable release. Patches after release will be indicated by adding third number.
We do lot of changes in system internals, and much more expected in next milestone versions. Although we do our best to test system before release, M versions may still be unstable, so upgrade with care and be prepared for rollback.

Full change log since previous release:

- New hook script "AcceptNewNode" (can be used as additional filter for network discovery to avoid unnecessary communications)
- SMS driver for Kannel gateway
- Oracle subagent: added parameter Oracle.CriticalStats.Deadlocks
- Reporting server improved
- Code page option added to all client command line tools (nxalarm, nxevent, nxpush, nxsms)
- New NXSL functions: GetNodeTemplates
- New parameters in Windows agent: System.Update.LastDetectTime, System.Update.LastDownloadTime, System.Update.LastInstallTime
- Network service response time can be monitored
- Added driver for HP A-series and V-series switches
- MIB compiler correctly handles REFERENCE keyword in agent capabilities section
- Oracle Tuxedo monitoring subagent
- Improved server performance with PostgreSQL
- Alias and MTU collected and stored for interfaces
- Server uses ipAddressTable and ipAddressPrefixTable to collect configured IP addresses when supported by monitored nodes
- Management console:
    - New dashboard element "DCI summary table"
    - Cluster objects shown on status maps
- Android console:
    - Added date in X axis (graphs) when the time frame is greather that one day
    - Added 10 minutes, 12 hours, 5 days, 30 days, last year options in draw graph for last values tab (5, 7, and 30 days asks for confirmation of long task execution)
    - Fixed graphical problem (checkbox) in last values tab
    - Support for KitKat Wear and Lollipop, updated support library v4
- Fixed issues: #61, #306, #381, #488, #565, #628, #634, #667, #672, #673, #674, #675, #677, #678, #679, #682, #684, #686, #688, #689, #691, #697, #698, #700, #704, #705, #709, #710, #715

Best regards,
Victor

Hi,

browsing performance counters from template configuration is not supported yet. You can configure required counter on node and then use "convert to template item" option to move it to template.

Best regards,
Victor