[OPNFLWPLUG-2] Stats manager sends request for stats to non-openflow nodes Created: 03/Dec/13  Updated: 27/Sep/21  Due: 20/Jan/14  Resolved: 25/Jan/14

Status: Resolved
Project: OpenFlowPlugin
Component/s: General
Affects Version/s: None
Fix Version/s: None

Type: Bug
Reporter: Jan Medved Assignee: Anil Vishnoi
Resolution: Done Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified
Environment:

Operating System: Mac OS
Platform: Macintosh


External issue ID: 205

 Description   

2 netconf nodes in inventory, no openflow nodes at all. I see the following in the logs:

2013-12-02 22:47:39.235 PST [Thread-21] INFO o.o.c.m.s.manager.StatisticsProvider - Send request for stats collection to node : Uri [_value=controller-server])
2013-12-02 22:47:39.235 PST [Thread-21] INFO o.o.c.m.s.manager.StatisticsProvider - Send request for stats collection to node : Uri [_value=yuma])
2013-12-02 22:48:29.236 PST [Thread-21] INFO o.o.c.m.s.manager.StatisticsProvider - Number of connected nodes : 2



 Comments   
Comment by Ed Warnicke [ 03/Dec/13 ]

Quick comment... StatisticsManager.statsRequestSender is
sending requests to all Nodes to known to the system... but it should only send messages to all known FlowCapableNodes... not all Nodes are FlowCapable Nodes

Comment by Anil Vishnoi [ 17/Jan/14 ]

Hi Jan,

I fixed this issue and pushed code changes through

http://git.opendaylight.org/gerrit/4356

Can you please test it once and confirm that it fixes the issue.

Will close this bug afterward.

Thanks
Anil

Comment by Jan Medved [ 21/Jan/14 ]

This is now blocking Netconf testing. Requesting the entire operational data tree from a large netconf node (e.g. xr) can take a long time.

Comment by Anil Vishnoi [ 21/Jan/14 ]

Hi Jan,

I already fixed this issue and the code changes are already merged to code base. I changed the code so that it only sends stats request to FlowCapableNode only.

Do you still see this issue with latest code base?

If yes, then there can be two possible reason:
1) Either netconf devices are FlowCapableNode, is it the case ??
2) or getAugmentation has some bug, and its passing following condition

if(targetNode.getAugmentation(FlowCapableNode.class) != null)

Thanks
Anil

Comment by Lukas Sedlak [ 25/Jan/14 ]

I just submitted fix for this issue, it's on https://git.opendaylight.org/gerrit/#/c/4764/ I added support for hexadecimal, octal and decimal numerical systems as are stated in RFC 6020.

There are still rules that users have to follow:

An integer value is lexically represented as an optional sign ("+" or "-"), followed by a sequence of decimal digits.

The hexadecimal notation consists of an optional sign ("+" or "-"), the characters "0x" followed a number of hexadecimal digits, where letters may be uppercase or lowercase.

The octal notation consists of an optional sign ("+" or "-"), the character "0" followed a number of octal digits.

More on http://tools.ietf.org/search/rfc6020#section-9.2.1

Although the yang specification clearly states that: "In the XML instance documents, an integer is always interpreted as a decimal number, and leading zeros are allowed." Thus in general users SHOULD use decimal numbers in xml instances.

Generated at Wed Feb 07 20:31:19 UTC 2024 using Jira 8.20.10#820010-sha1:ace47f9899e9ee25d7157d59aa17ab06aee30d3d.