Problem with creating new Network Maps [SOLVED]

Started by sperlm, August 15, 2013, 12:25:07 PM

Previous topic - Next topic

sperlm

Hello, I am struggling with creating Network Maps...

After upgrading to 1.2.8 every time I try to create any type of Network Map I get this "Problem Occured - 'Create new network map' has encountered a problem. Cannot create network map object "test": Request timed out".
Not sure if the problem appeared with new version as the map adding feature was needed only recently.

Already created maps work fine.

With regards

Milan Sperl

Victor Kirhenshtein

Hi!

Can you enable debug on server and check server log when you create map object? I cannot reproduce this on test system, so it should be related to your specific system and/or configuration.

Best regards,
Victor

sperlm

#2
Hello, back after some debugging-fixing...

Thanks for pointing the debug mode, level 2 found some obvious errors which I managed to fix.
Tried to troubleshoot level 3 after that, not succesfully for now...
The main problem still remains with one exception - managed to create one IP topology map (for one locality only).

In debug level 3 there are obvious parts which might cause the problem I have:
[03-Sep-2013 13:51:53.837] NetworkMap::updateContent(Cerny Most [8770]): call to buildL2Topology on object 5008 failed
[03-Sep-2013 13:51:53.839] NetworkMap::updateContent(Krc [21794]): call to buildL2Topology on object 5181 failed
[03-Sep-2013 13:51:53.839] NetworkMap::updateContent(Jizni Mesto II [21874]): call to buildL2Topology on object 6282 failed
[03-Sep-2013 13:51:53.839] NetworkMap::updateContent(Neratovice [23851]): call to buildL2Topology on object 22855 failed
[03-Sep-2013 13:51:53.839] NetworkMap::updateContent(Lehovec [32983]): call to buildL2Topology on object 3975 failed
[03-Sep-2013 13:51:53.847] NetworkMap::updateContent(Kukelska [36863]): call to buildL2Topology on object 12123 failed

These are mostly IDs of Network Maps and IDs of the base object.

There were some warnings during initialization too, which I would like to fix but failed for now:
[03-Sep-2013 13:50:50.350] WARNING: duplicate interface IP address 0AFEFE09 (interface object Vlan-interface212 [22311])
[03-Sep-2013 13:50:50.615] WARNING: duplicate interface IP address 0AFEFEED (interface object Vl513 [22530])
[03-Sep-2013 13:50:53.217] WARNING: duplicate interface IP address 0AFEFE75 (interface object Vlan-interface108 [36849])
[03-Sep-2013 13:50:53.219] WARNING: duplicate interface IP address 0A006001 (interface object Vlan-interface800 [36851])
[03-Sep-2013 13:50:53.221] WARNING: duplicate interface IP address 0AF89D01 (interface object Vlan-interface801 [36852])
[03-Sep-2013 13:50:53.223] WARNING: duplicate interface IP address 0AFEFE1E (interface object Vlan-interface206 [36853])
[03-Sep-2013 13:50:53.229] WARNING: duplicate interface IP address AC1064C8 (interface object Vl510 [36856])
[03-Sep-2013 13:50:53.231] WARNING: duplicate interface IP address 0AFEFE69 (interface object Vlan-interface501 [36857])
[03-Sep-2013 13:50:53.233] WARNING: duplicate interface IP address 0A005B58 (interface object Vlan-interface91 [36870])

I would like some pointers how to fix these.
Tried to search in database for nodes where these ports are but not sure how to fix it (already tried removing the ports and polling configuration afterward).

Attaching the whole level 3 log to be sure.
p.s. I suppose the errors at the scripts section are just warnings, not interfering with anything (I know these scripts are not fixed/finished)

Victor Kirhenshtein

Regarding duplicate IPs: check in console if these interfaces really have same IP addresses as some other interfaces.

Can you please run topology poll from console (Poll -> Topology from context menu on node) on any of switches where error "call to buildL2Topology on object 5008 failed" displayed and post output here?

Best regards,
Victor

sperlm

Hello,

* regarding network maps
- attaching 2 topology polls, nothing really supsicious though...

* ragarding duplicate IPs
had to look in SQL database and dig for:
- "duplicate" interface ID
- then for the IP found in the interface ID
Then look in NetXMS console and:
- search Interface view for the ports
- check by Find IP function

Results:
- in all cases both interfaces are shown in Interfaces tab in NetXMS for both devices
- interface with the ID in the error log is the one which NetXMS Find IP function finds

Should I delete and recreate the other interface then ?

Quote from: EXAMPLE[11-Sep-2013 10:08:21.100] WARNING: duplicate interface IP address 0AFEFE09 (interface object Vlan-interface212 [22311])
--SQL dig on interface 22311
22311, 3881, 0, '10.254.254.9', '255.255.255.252', 136, 37, 0, 0, 0, 0, 0, 'D07E28B6C3FC', 0, 1, 1, 0, 0, 'Vlan-interface212'
--SQL dig on interface-IP 10.254.254.9
3809, 3801, 0, '10.254.254.9', '255.255.255.252', 136, 855, 0, 0, 0, 0, 0, '001AC19AD001', 0, 1, 2, 0, 0, 'Vlan-interface104'
22311, 3881, 0, '10.254.254.9', '255.255.255.252', 136, 37, 0, 0, 0, 0, 0, 'D07E28B6C3FC', 0, 1, 1, 0, 0, 'Vlan-interface212'
--NetXMS Console Find IP 10.254.254.9
Rack_Vysocanska_POP   Vlan-interface212   D0:7E:28:B6:C3:FC   10.254.254.9   Cakovice-POP   GigabitEthernet1/0/25   indirect
(this is the device 3881 with interface 22311)

* both interfaces are shown in the Interfaces tab in NetXMS.

With regards

Milan Sperl

Victor Kirhenshtein

Hi!

And what interface is correctly discovered - Vlan-interface104 or Vlan-interface212? Or they both really have same IP address? If not, just delete wrong interface (or  both - system will re-create correct one on next configuration poll).

I didn't see anything wrong with topology polls. I'll add more debug output for map generation in upcoming release - hopefully it will help identify why maps are not created as expected.

Best regards,
Victor

sperlm

#6
As I wrote - both interfaces are shown in Interfaces tab in NetXMS for both devices
Code (Interface for device 3801) Select
3809 Vlan-interface104 136 855 Vlan-interface104 00:1A:C1:9A:D0:01 10.254.254.9 UP DOWN UP Unmanaged
Code (Interface for device 3881) Select
22311 Vlan-interface212 136 37 Vlan-interface212 D0:7E:28:B6:C3:FC 10.254.254.9 UP UP UP Normal

Find IP function returns one row only (device ID 3881, with interface ID 22311)
1 Rack_Vysocanska_POP Vlan-interface212 D0:7E:28:B6:C3:FC 10.254.254.9 Cakovice-POP GigabitEthernet1/0/25 indirect
The same applies for all other cases.

Deleted and ran configuration poll for both interfaces - both recreated the same way, only with different ID.
Anyway, this kind of "error" which might be intentional configuration cannot be the cause of the map creation problem, right?

Will do some more map testing, I still think it might be node related rather than general.
I thought about using higher debug level than 4 if there were no messages about the problem. Its just that it creates rather big log very fast... do you have any suggestions what level should I try or what to look for?

On a side note - had some issues with space for the DB (can this be the cause of network map creation timeouts?) and eventually run out of space completely. It seems that creating automated DCIs (for 30 day period) might not be a wise move as every object takes approximately 25 MB of space in MySQL database. So we are planning to backup-restore the database to save up space after limiting the DCI spread (and duration to 7 days), for the time being.

With regards

Milan Sperl

sperlm

Hi,

   just a quick heads up - the whole problem seems to originate from insufficient disk space.
Looks like playing with tons of DCIs to allocate great deal of free space was the main problem.
After realizing state of the free disk space left we toned down the automatically created DCI amount, backup-restored sql. After that map creation was functional again.

With regards

Milan Sperl

Victor Kirhenshtein

Hi!

That's good to hear that problem source is found - problem seemed very weird.

Best regards,
Victor