NetXMS Support Forum

Hi,
We are having some problems with 2.2.7.  nxdbmgr reports clean on a "check" and "check_data_tables".
On Startup things start up it seems ok but throws an error:


2018.07.22 12:48:57.007 *D* 28 network device drivers loaded
2018.07.22 12:48:57.007 *D* Built-in objects created
2018.07.22 12:48:57.007 *D* SQLite version 3.22.0
2018.07.22 12:48:57.007 *I* Database driver "sqlite.ddr" loaded and initialized successfully
2018.07.22 12:48:57.007 *D* [db.conn            ] New DB connection opened: handle=000000FE67A8E630
2018.07.22 12:48:57.007 *D* Caching object configuration tables
2018.07.22 12:49:17.694 *D* Loading built-in object properties...
2018.07.22 12:49:17.710 *D* Loading zones...
2018.07.22 12:49:17.710 *D* NetObj::loadCommonProperties() failed for object Default [4] class=6
2018.07.22 12:49:17.819 *D* Loading conditions...

It continues on and loads everything else ... starts polling and tunnel creation and if I check the show q and show thr  in nxadm everything looks ok.  Show watchdog responds:

netxmsd: sho w
Thread                                           Interval Status
----------------------------------------------------------------------------
Item Poller                                      10       Running
Syncer Thread                                    30       Sleeping
Poll Manager                                     5        Sleeping
Ad hoc scheduler                                 5        Sleeping
Recurrent scheduler                              5        Sleeping

A couple of minutes later polling seems to stop (with a lot of active requests waiting) and the watchdog does not respond any more.

netxmsd: sho thr
MAIN
   Threads:            42 (8/1024)
   Load average:       0.32 0.12 0.04
   Current load:       0%
   Usage:              4%
   Active requests:    0
   Scheduled requests: 0

POLLERS
   Threads:            4096 (2048/4096)
   Load average:       10964.91 3197.29 1102.58
   Current load:       414%
   Usage:              100%
   Active requests:    16967
   Scheduled requests: 0

DATACOLL
   Threads:            1024 (10/1024)
   Load average:       1220.03 569.57 211.84
   Current load:       2%
   Usage:              100%
   Active requests:    28
   Scheduled requests: 0

SCHEDULER
   Threads:            1 (1/64)
   Load average:       0.00 0.00 0.00
   Current load:       0%
   Usage:              1%
   Active requests:    0
   Scheduled requests: 0

AGENT
   Threads:            61 (4/512)
   Load average:       12.35 2.86 0.94
   Current load:       100%
   Usage:              11%
   Active requests:    61
   Scheduled requests: 0

SYNCER
   Threads:            1 (1/10)
   Load average:       0.00 0.00 0.00
   Current load:       0%
   Usage:              10%
   Active requests:    0
   Scheduled requests: 0

netxmsd: sho w
Thread                                           Interval Status
----------------------------------------------------------------------------



I can leave this for hours and nothing changes.  The only activities are tunnels opening and closing and threads stopping in data collection due to inactivity.

The only other error message that I can find in the log files is:

Cannot establish connection with ISC peer ::1

I have restarted the netxmsd service many times.
I have rebooted multiple times.
Windows updates are all up to date.  (Windows Server 2012r2 running in a hyperv vm, 32GB 10CPU's)
Database is on a separate 2012 server with MSSQL 2014 with lots of space.

Any help would be appreciated

Thx.

I maybe a little thick......would i create a new script called Hook::DiscoveryPoll or add that into one of the existing Hook:: scripts?

I currently have:
11 - Hook::StatusPoil
12 - Hook::ConfigurationPoll
13 - Hook::InstancePoll
15 - Hook::CreateInterface
16 - Hook::AcceptNewNode

Manually Starting a Discovery Poll on a subnet. 

As far as existing hook scripts ...I don't see one that would have access to the subnet object (at least not in the comments).

Since the subnet doesn't show up unless there is a node in it, that means there is a node to do the discovery on.  I assume it then looks at the arp table , fdb, cdp, or lldp,  of the existing nodes for new nodes. 

As to when to do the hook, either at the beginning of the process or the end, i guess it doesn't matter.  what i would be looking to do is set a custom attribute with a last run date/time, whether or not it found anything new.

We also have this problem.....resolving or terminating as well as other functions on nodes or containers (renaming, binding, etc).  The timeout dialog is displayed but the operation DOES happen.  It just takes some time to display in the management console. 

would it be possible to add a manual discovery poll on a subnet? similar to a manual node poll?
i.e. (EntireNetwork->(zoneUIN)->subnet/mask) rightclick -> poll -> discovery

Along with a Hook::DiscoveryPoll script that could be called even if no new nodes are added.  This way a subnet-customattribute could be updated with a last run date/time similar to Alex's configuration poll example.

Thx.

Yes you can....here is an example:

WMI.Query(root\cimv2, "select * from SoftwareLicensingProduct where partialproductkey like '%'" , LicenseFamily )

Yeah I tried wildcards too.....didn't work.
Ugly, but I ended up with a if( ){} structure for each interface name that i didn't want and that worked.
(elminated about 9600 interfaces)

To limit the (re)addition of unwanted interfaces there is a script in the scriptlibrary called Hook::CreateInterface

There are some examples on the site here, but simply you can do:

if ($1->name ilike "Po1"){
   return false;
   }
return true;

Replace Po1 with the interface name that your switch is reporting.

My 2 cents:

There are a number of ways to do this currently.
You can setup containers to group nodes any way you want (or multiple ways).  I have an infrastructure tree with around 25000 containers that group subsets of 5500 nodes in 100's of different ways.

You could also setup "custom attributes" on the nodes in question and then create a seldom polled dci that queries the custom attribute and have it show up in the "last values" panel on the overview tab of the object.

thx

I had not seen this parameter before and my googlefu seems to off at the moment.

Any chance that this will be a standard agent parameter again for win 10 or server2016 ?

With the New agent there are parameters called:
System.Update.LastInstallTime
System.Update.Lastdownloadtime
System.Update.lastdetect

with windows 10 and windows server 2016 they are throwing the following error in the agent log file:

GetParameterValue(): result is 500 (INTERNAL ERROR)

I don't have any 32bit windows 10 machines to test this however it does work on Windows 7 and Windows Server 2012r2

Yes, Thank you very much for your great work....

p.s. Why the number shift (2.2.1 vs 2.1.3)?

In server console or nxadm -i,
show stat

This will show # of objects, nodes , and DCI's

I know I'm lazy but it would be nice to be able to clone existing DCI Summary Tables.

thx