NetXMS Support Forum

Quote from: weec on February 07, 2008, 09:25:43 AM
а какже возможность биндить сервисы на определенные интерфейсы?

Все будет Это еще не релиз.

Теперь понял. Сделаю.

Quote from: Alex on February 06, 2008, 10:51:41 PM
Вопрос 1:
При установленном значении: "2" если я сделаю какие-то изменения в Properties, они останутся или затрутся при Poll->Configuration?

Zatrutsja. Ja sdelaju takze vozmoznost' menjat'imja interfeisa i ne zatirat' ego pri konfiguration poll'e, no pozze. Voobsche mne eto kazetsja ideologicheski ne ochen' pravil'nim - po idee dlja custom informacii nado ispol'zovat' kommentarii k ob'ektu. No u menja samogo setevih ustrojstv malo, tak chto bol'shogo prakticheskogo opita net - ne znaju kak v real life udobnee.

Quote from: Alex on February 06, 2008, 10:51:41 PM
Вопрос 2:
Можно ли добавить дополнительную фишку помимо Poll-->Configuration и Poll-->Status еще одну Poll-->Descriptions???

Tehnicheski konechno mozno. A chem ploh configuration poll?

Quote from: Alex on February 06, 2008, 10:51:41 PM
Вопрос 3:
Есть какие-то еще новшества в этой версии?

Tekushij change log po otnosheniju k 0.2.19:

- Implemented "proxy node" functionality for DCIs
- Added possibility to specify multiple recipients in e-mail or SMS action
- Implemented System.CPU.Usage for individual processors on Linux
- PING subagent: added parameter ICMP.PacketLoss(*)
- Command line options for the server changed to use common style
- Added UNICODE support in client part on Linux/UNIX
- Implemented macros in template DCIs (expanded when template applies to node)
- Added event storm detection
- Added possibility to specify multiple SNMP community strings for discovery
  and configuration polls
- NetXMS server now can understand interface aliases (description in Cisco terms)
- Added optional synchronization of node names with DNS
- Fixed issues: #189, #190, #199

Можно скачивать по адресу https://www.netxms.org/download/rc

Implemented in 0.2.20. This feature is disabled by default, to enable it you should set server configuration variable SyncNodeNamesWithDNS to 1.

Best regards,
Victor

Just check your spreadsheet - excellent! Can I publish it on web site in documentation section as well?

Best regards,
Victor

Скоро должна выйти. Если хочется, я могу сегодня собрать и выложить release candidate. Заодно и бета-тестеры появятся...

В версии 0.2.20 реализована поддержка определенных пользователем названия интерфейсов (interface aliases). Использование контролируется параметром сервера UseInterfaceAliases. Могут быть следующие значения:
0 - не использовать
1 - в качестве имени интерфейса использовать алиас
2 - конкатенировать алиас и имя интерфейса

Currently, the only thing you can easily monitor is interface up/down. System detect all interface state changes and send predefined events: SYS_IF_DOWN and SYS_IF_UP. You can filter and process these events in event processing policy, so you will need to configure only a few rules in EPP.
For other things, there are some work to do

Best regards,
Victor

Quote from: rmz on February 06, 2008, 05:10:30 AM
One more quick question.  Using the object map, I can watch the various managed nodes for their status.  In the case of an unmanaged node is it possible to some how use this data collection point (icmp.ping values) with an unmanaged node on the object map, so I can verify status for them as well? 

You can just switch them to managed - if node doesn't have neither SNMP agent nor NetXMS agent, server will just ping it regularly and change status accordingly. Alternatively, you can define DCI for ICMP.ping(unmanaged_node_adderess) parameter on node with PING subagent loaded. But in this case, status of the unmanaged node object will not be changed.

Best regards,
Victor

Actually I create one table per node, not per DCI.

Best regards,
Victor

Try to stop IPSEC Services if they are running and are not needed. I can't remember exactly, but probably I've seen that running IPSEC Services interfere with custom programs working with ICMP while system ping works fine. Anyway, I'll try to find more information.

Best regards,
Victor

Thank you for spreadshit. I'll test it tomorrow, when I'll have MS Office at hands.

Quote from: kghammond on February 05, 2008, 07:00:24 PM
Also on a side note, are you creating a clustered index or any non-clustered indexes on the tables that store the rows of data?

For each table there are one clustered and one non-clustered index.

Best regards,
Victor

Also, Microsoft has document for estimating table size:

http://msdn2.microsoft.com/en-us/library/aa933068(SQL.80).aspx

NetXMS stores collected data in individual tables for each node, and each table has three fields: two integer and one varchar(255). However, if you collect mostly numerical data, you can assume that varchar field is used for about 10% of it's full capacity.

Best regards,
Victor

All collected data stored in varchars, as a textual representation of appropriate data type. I don't now how much space takes each record in database, but can provide some real life statistics.

My own NetXMS server runs on MS SQL 2005, I have 101 node with 1769 DCIs configured, almost all of them configured for collection every 60 seconds and retention time of 30 days (some configured for 3600 seconds and 365 days). Database size now is around 6.5 GB. For your situation, you will have about 400 DCIs for 100 switch interfaces and about 300 to 500 (depending on number of partitions, required parameters, etc.) for 40 servers - so probably your database size for 60 seconds/30 days configuration will be about 4GB in size.

I will try to collect data for some other MS SQL installations and post them here.

Best regards,
Victor