Bug 1818

Summary: FlowMonitor needs IPv6 support
Product: ns-3 Reporter: Tommaso Pecorella <tommaso.pecorella>
Component: flow-monitorAssignee: Gustavo J. A. M. Carneiro <gjcarneiro>
Status: RESOLVED FIXED    
Severity: enhancement CC: ns-bugs, tomh
Priority: P5    
Version: ns-3-dev   
Hardware: All   
OS: All   
Bug Depends on: 1808, 1843    
Bug Blocks:    
Attachments: patch
new manual
new patch
updated manual

Description Tommaso Pecorella 2013-12-24 10:45:59 EST
Created attachment 1747 [details]
patch

Obviously IPv6 support is needed by FlowMonitor.
Comment 1 Tom Henderson 2014-01-23 10:32:18 EST
I scanned the patch and had one question and one comment.

Why is this commented out?
+//        case Ipv6L3Protocol::DROP_FRAGMENT_TIMEOUT:
+//          myReason = DROP_FRAGMENT_TIMEOUT;
+//          NS_LOG_DEBUG ("DROP_FRAGMENT_TIMEOUT");
+//          break;

The comment is that it would be preferable to have some test code to exercise this, even though the original (IPv4 case) lacks test code coverage.
Comment 2 Tommaso Pecorella 2014-01-23 13:38:50 EST
Created attachment 1767 [details]
new manual

This is the FlowMonitor updated manual. Gustavo already checked it.
Comment 3 Gustavo J. A. M. Carneiro 2014-01-23 13:44:54 EST
Something I missed before: there are a few unanswered questions in the text:

"What kind of data does the model generate?  What are the key trace
sources?   What kind of logging output can be enabled?"

Some copy-pasted template?
Comment 4 Tommaso Pecorella 2014-01-23 13:47:14 EST
(In reply to Tom Henderson from comment #1)
> I scanned the patch and had one question and one comment.
> 
> Why is this commented out?
> +//        case Ipv6L3Protocol::DROP_FRAGMENT_TIMEOUT:
> +//          myReason = DROP_FRAGMENT_TIMEOUT;
> +//          NS_LOG_DEBUG ("DROP_FRAGMENT_TIMEOUT");
> +//          break;

The answer is: it's [missing feature] in IPv6. When a packet is lost due to fragment timeouts, the drop trace is not called.
I'll fix it (in another bug). before applying this one, of course.

> The comment is that it would be preferable to have some test code to
> exercise this, even though the original (IPv4 case) lacks test code coverage.

Yeah… it would :P
As a matter of fact, FlowMonitor lacks tests more or less for everything.
However, if you don't mind, I'd fix that later on.
Comment 5 Tommaso Pecorella 2014-01-23 13:53:09 EST
(In reply to Gustavo J. A. M. Carneiro from comment #3)
> Something I missed before: there are a few unanswered questions in the text:
> 
> "What kind of data does the model generate?  What are the key trace
> sources?   What kind of logging output can be enabled?"
> 
> Some copy-pasted template?

Yes, I forgot those sections. I'll add them.
Comment 6 Tommaso Pecorella 2014-01-24 03:30:23 EST
Created attachment 1769 [details]
new patch
Comment 7 Tommaso Pecorella 2014-01-25 07:41:54 EST
Created attachment 1771 [details]
updated manual

New manual version. Clarified some points, added an XML example and removed unused sections.
Comment 8 Tommaso Pecorella 2014-03-17 09:47:09 EDT
Comment on attachment 1771 [details]
updated manual

Moved to bug #1802
Comment 9 Tommaso Pecorella 2014-04-21 05:37:21 EDT
pushed in changeset 10692:bdf59f5ace93