NetXMS Support Forum

Hi,

most likely you already have (presumably incorrect) subnet object for /24. You can check what's inside and delete subnet object (make sure that nodes inside it bound to at least one container, then they will not be deleted with the subnet), then do configuration poll on a node - it should create correct /26 subnet object.
Also, in 2.0 we do some fixes to subnet creation and binding logic, it will help in such situations.

Best regards,
Victor

Про world map - согласен, и сделать это по идее не сложно. Добавил в трекер (https://dev.raden.solutions/issues/777).

Про базы не совсем понял - какие именно базы и как их объединять?

Сейчас можно настроить что-то похожее. Для любой ноды можно создать internal DCI PingTime. Соответственно в Last Values будет время отклика на ping во время последнего status poll. Затем можно настроить DCI Summary Table по этому параметру и видеть в виде таблицы времена отклика для узлов.

Do you have WAR products feature (org.eclipse.libra.warproducts.feature.feature.group) installed? If yes, what version? I use 0.3.1.

Best regards,
Victor

Hi,

O_RDONLY means that agent wants read only access to the file. Sharing mode is controlled by third argument, which is set to _SH_DENYNO, meaning that agent allows both read and write shared access to the file. However, the problem may be with the application writing to the file - if application requests exclusive access (i.e. not permitting read access by others), it will not be able to open file for writing. We cannot do anything about it. If you can control the application in question you can change it or request application developers to allow shared read access to the log.

Best regards,
Victor

Hi,

we are gradually implementing IPv6 support under the hood. Management console and agents can communicate over IPv6 already, and there are internal changes to support both IPv4 and IPv6 addresses on objects. Final 2.0 release will fully support IPv6, and you will see IPv6 subnets in M3 or M4.

Best regards,
Victor

Hi,

no, currently it's not possible. Implementing such feature within NetXMS itself can be quite complicated because we will need to implement in-memory graph creation on server side (currently all visualization done on client side).

Best regards,
Victor

Poll count 300 means 5 hours to change status - is it what you intend?

Best regards,
Victor

I've added session agent termination to  agent installer. The only problem is that session agent will be unavailable until restart because installer will not be able to start it again in another user session.

Best regards,
Victor

If those nodes respond to ICMP ping, you can try to configure active discovery for problematic networks. Also, NetXMS will not add non-responding nodes - are they accessible from NetXMS server?

Best regards,
Victor

Hi,

please send me output of query

SELECT * FROM metadata

from your database. For now, try to answer y only to invalid object fix - that way it will fix at least this problem.

Best regards,
Victor

Hi,

you can change required poll count for node interfaces. For example, if you set it to 5, server will consider interface as down if 5 consecutive polls will report that. Node will not be marked as down as long as interfaces not marked as down.

Best regards,
Victor

В багтрекере лучше все писать по английски.

Вытянуть кусок файла в DCI - только через ExternalParameter (еще можно написать скрипт как ExternalParametersProvider если надо сразу много строк в виде параметров из файла вытащить).

Багтрекер находится здесь: https://dev.raden.solutions.

Для просмотра файлов можно использовать filemgr субагент, для мониторинга логов - logwatch. Доступа к registry сейчас нет, но в 2.0-M3 добавим (уже давно висит запрос на это).

сейчас это возможно только через ping субагент. Для всего, что настроено как target считается packet loss (доступен через параметр Icmp.PacketLoss).