NetXMS Support Forum

Hi!

Bug should be fixed. We have uploaded updated installers, Please download updated version and reinstall.

Best regards,
Victor

Hi all!

NetXMS version 1.2.10 is released. Changes from previous release:

- Event processing policy rules can be exported and imported
- NXSL:
   - Implemented post-increment and post-decrement for array elements
   - New functions: GetDCIValues
- Server's ICMP ping timeout can be configured
- DB/2 monitoring subagent
- DB/2 supported as backend database
- Table DCIs can be used in condition objects
- nxapush can push data on behalf of other nodes
- Transformation script can be run on aggregated DCIs
- Unknown SNMP trap sources can be automatically added as nodes
- Syslog processing optinmized for performance
- Default shared secret for agents can be configured
- Driver for Ping3 devices
- Driver for Allied Telesis switches
- Agent installer on Windows automatically adds firewall exception
- Database migration tool
- Additional parameters in Solaris platform sub-agent
- Management console:
   - New command line option -fullscreen
   - Option to draw borders around each value in "gauge" dashboard elements
   - Sorting by value in "Last Values" view takes data type into consideration
   - "Nodes" tab in object details view for subnets and containers
   - "Address Map" tab in object details view for subnets
   - Czech and Russian localization
- Fixed issues: #150, #226, #307, #330, #362, #363, #364, #366, #367, #373, #378, #384, #387, #390, $401, #410, #414, #416, #419, #439, #440

Best regards,
Victor

Hi!

It's not possible yet. All you can do is to setup two thresholds and use filtering script in event processing policy to block events at specific time of the day. However, soon (I plan for 1.2.11 so far) you'll be able to create thresholds as NXSL scripts, then it will be much easier.

Best regards,
Victor

Hi!

I've changed the drbd.cpp (protocol version should also work as expected). These changes are included in 1.2.10 release. Thank you for fixing it!

Best regards,
Victor

*PING - это начало новой секции конфига. ActionShellExec должен быть в основной секции, вот так:


MasterServers = trampampam.ru
ConfigIncludeDir = C:\NetXMS\etc\nxagentd.conf.d
logFile = C:\NetXMS\var\log.txt
FileStore = C:\NetXMS\var
SubAgent = ecs.nsm
SubAgent = ping.nsm
SubAgent = logwatch.nsm
SubAgent = portcheck.nsm
SubAgent = winperf.nsm
SubAgent = wmi.nsm
SubAgent = ups.nsm

ActionShellExec = ResroreWinVPN:C:\Windows\system32\rasdial.exe $1 $2 $3

*PING
Timeout = 1000
PacketRate = 12
Target = 10.0.0.1:target_1


Hi!

1.2.10 should be released today.

Best regards,
Victor

Hi!

You can create new internal DCI with custom schedule to run once per month and either read current value of cumulative counter in transformation script. The tricky part is timing to ensure that monthly DCI runs before cumulative counter resets at month change. Alternative option could be storing value temporarily in node's custom parameters - when cumulative counter resets, it stores it's current value into some custom attribute, and then monthly DCI reads it.

Best regards,
Victor

You returning value of original DCI in your transformation script, but if you want cumulative counter, you should return sum of current value of your dummy DCI and current value of original DCI, something like

GetDCIValue($node, dci->id) + GetDCIValue($node, $dci->id)

($dci refers to current DCI being processed).

And to reset the counter - check if last poll time for your dummy DCI is in different month, and return 0 in that case, like this:

if (strftime("%m", $dci->lastPollTime) < strftime("%m", time()))
   return 0;

Best regards,
Victor

This filtering script checks that absolute value for given OID is 0. If you have non-zero value which is not changing over time, OID will pass filter but you will still get 0 as delta value. If you want to filter out any non-changing values, you should get two samples and check delta, like this:


snmp = CreateSNMPTransport($node);
v1 = SNMPGetValue(snmp, ".1.3.6.1.2.1.2.2.1.16." . $1);
sleep(10000);  // sleep for 10 seconds
v2 = SNMPGetValue(snmp, ".1.3.6.1.2.1.2.2.1.16." . $1);
return (v2 - v1) != 0;


Of course, practical use of such method is questionable and depends on your goals - you will either miss slowly changing OIDs if sleep interval is too small or make configuration polling extremely long if you have lot of instances and long sleep interval.

Best regards,
Victor

Hi!

This could happen if you have large network, autodiscovery is on, and lot of hosts/devices did not respond to ping and SNMP (usually because of firewall). If this is the case, you don't have to worry - eventually queue will go down. The only drawback is that network discovery will take much longer type. You can speed up discovery process by allowing access from NetXMS server where needed and/or using address filters for network discovery to skip unneded address ranges.

Best regards,
Victor

Hi!

SNMP OID used for data collection is in fact DCI name, and so it is available as first event parameter via macro %1 or variable $1 in filtering scripts.

Best regards,
Victor

Hi!

I've changed sorting in last values view to be data type aware. This fix will be part of 1.2.10 release.

Best regards,
Victor

Вот он.

Почти все выложено. Не переведен из выложенного ObjectManager. Ну и чуть позже подготовлю PerfView - и это будет последний плагин

С неравными интервалами могут быть только push параметры, но тогда придется сбор данных выносить во внешний скрипт. Можно сделать по другому - в DCI хранить кол-во тонера, и поставить threshold на условие diff with previous value > 0 - тогда он сработает когда кол-во тонера вырастет, т.е. картридж был заменен. По событию в этот момент можно через SNMP получить кол-во страниц (как сейчас сделано в transformation script), и использовать его в аларме и/или оповещениях. Дополнительный бонус - можно вывести на график текущий уровень тонера.