feat: request UPF to allocate F-TEID #334

gruyaume · 2024-10-16T19:57:34Z

Description

As of 3GPP release 16, the F-TEID should be allocated by the UPF (and not by the SMF). Here we make the necessary changes for the SMF to request the UPF to allocate the F-TEID through PFCP and we remove the F-TEID allocation from the SMF.

⚠️ Warning: This is a breaking change. This PR goes hand in hand with the change in the UPF.

Context

From the 3GPP specification referenced below:

For EPC and 5GC, F-TEID shall be only allocated by the UP function, see clause 5.8.2.3 of 3GPP TS 23.501[28].
The UP function shall set the FTUP feature flag in the UP Function Features IE (see clause 8.2.25) and the CP function
shall request the UP function to allocate the F-TEID. The UP function shall reject a request to establish a new PDR with
an F-TEID allocation in the CP function option, with the cause "Invalid F-TEID allocation option". As an exception, the
UP Function shall accept the PFCP Session Establishment Request message with a PDR including an existing F-TEID
to re-establish a PFCP session during a restoration procedure as specified in clause 4.3.2 of 3GPP TS 23.527 [40] and
clauses 16.1A.4 and 17.1A.4 of 3GPP TS 23.007 [24].

Reference

Signed-off-by: Guillaume Belanger <[email protected]>

thakurajayL · 2024-10-19T10:47:24Z

Not reviewing since its in draft state !

Signed-off-by: Guillaume Belanger <[email protected]>

VERSION

Signed-off-by: Guillaume Belanger <[email protected]>

gab-arrobo

Overall, everything looks good

context/datapath_test.go

context/upf.go

Signed-off-by: Guillaume Belanger <[email protected]>

gruyaume · 2024-11-15T20:20:40Z

Overall, everything looks good

I believe I made the proper changes in the upf pfcp adapter here but I'm not very confident since I have no environment to validate this and the tests for this part of the code are non-existent.

gab-arrobo · 2024-11-16T00:06:18Z

Overall, everything looks good

I believe I made the proper changes in the upf pfcp adapter here but I'm not very confident since I have no environment to validate this and the tests for this part of the code are non-existent.

I am not sure I understand this comment. You mentioned that you have made the proper changes in the upf adapter, but, I do not see any PR open in that repo (https://github.com/omec-project/upfadapter/pulls). Or are you referring to the changes in smf/pfcp/adapter.go? If so, the changes did not work when running a test that includes/deploys the upfadapter pod

context/datapath.go

pfcp/adapter/adapter.go

pfcp/handler/handler.go

gruyaume · 2024-11-18T13:17:47Z

Overall, everything looks good

I believe I made the proper changes in the upf pfcp adapter here but I'm not very confident since I have no environment to validate this and the tests for this part of the code are non-existent.

I am not sure I understand this comment. You mentioned that you have made the proper changes in the upf adapter, but, I do not see any PR open in that repo (https://github.com/omec-project/upfadapter/pulls). Or are you referring to the changes in smf/pfcp/adapter.go? If so, the changes did not work when running a test that includes/deploys the upfadapter pod

I was referring to the changes in this commit: 1f33efe

I haven't changed upfadapter indeed, I can do my best to make the changes over there but I have not proper way of testing them.

Signed-off-by: Guillaume Belanger <[email protected]>

gruyaume · 2024-11-18T18:26:39Z

Overall, everything looks good

I believe I made the proper changes in the upf pfcp adapter here but I'm not very confident since I have no environment to validate this and the tests for this part of the code are non-existent.

I am not sure I understand this comment. You mentioned that you have made the proper changes in the upf adapter, but, I do not see any PR open in that repo (https://github.com/omec-project/upfadapter/pulls). Or are you referring to the changes in smf/pfcp/adapter.go? If so, the changes did not work when running a test that includes/deploys the upfadapter pod

I was referring to the changes in this commit: 1f33efe

I haven't changed upfadapter indeed, I can do my best to make the changes over there but I have not proper way of testing them.

I just looked at the upfadapter code and it's quite unclear to me what needs to be changed.

Note: This project has no README.md, which doesn't help in understanding its purpose.

go.mod

gruyaume · 2024-11-20T19:46:07Z

Overall, everything looks good

I believe I made the proper changes in the upf pfcp adapter here but I'm not very confident since I have no environment to validate this and the tests for this part of the code are non-existent.

I am not sure I understand this comment. You mentioned that you have made the proper changes in the upf adapter, but, I do not see any PR open in that repo (https://github.com/omec-project/upfadapter/pulls). Or are you referring to the changes in smf/pfcp/adapter.go? If so, the changes did not work when running a test that includes/deploys the upfadapter pod

I was referring to the changes in this commit: 1f33efe
I haven't changed upfadapter indeed, I can do my best to make the changes over there but I have not proper way of testing them.

I just looked at the upfadapter code and it's quite unclear to me what needs to be changed.

Note: This project has no README.md, which doesn't help in understanding its purpose.

@gab-arrobo were you able to re-test with the changes in the smf's pfcp adapter changes. I just looked at upfadapter and I can't find anything in there that would break the e2e tests. I believe the problem was in the SMF's pfcp adapter code. In any case, logs and/or captures would be necessary at this point.

gab-arrobo · 2024-11-20T19:49:12Z

@gab-arrobo were you able to re-test with the changes in the smf's pfcp adapter changes. I just looked at upfadapter and I can't find anything in there that would break the e2e tests. I believe the problem was in the SMF's pfcp adapter code. In any case, logs and/or captures would be necessary at this point.

I think I did run a test, but it did not work. Let me run another test this afternoon and I can provide logs.

gab-arrobo · 2024-11-21T02:51:35Z

@gab-arrobo were you able to re-test with the changes in the smf's pfcp adapter changes. I just looked at upfadapter and I can't find anything in there that would break the e2e tests. I believe the problem was in the SMF's pfcp adapter code. In any case, logs and/or captures would be necessary at this point.

I think I did run a test, but it did not work. Let me run another test this afternoon and I can provide logs.

@gruyaume the PRs are now working as expected when using the upfadapter

@thakurajayL, were all of your comments addressed? There is no more comments from my side

        "Profile Name: profile1, Profile Type: register",
        "Ue's Passed: 5, Ue's Failed: 0",
        "Profile Status: PASS",
        "Profile Name: profile2, Profile Type: pdusessest",
        "Ue's Passed: 5, Ue's Failed: 0",
        "Profile Status: PASS",
        "Profile Name: profile3, Profile Type: anrelease",
        "Ue's Passed: 5, Ue's Failed: 0",
        "Profile Status: PASS",
        "Profile Name: profile5, Profile Type: deregister",
        "Ue's Passed: 5, Ue's Failed: 0",
        "Profile Status: PASS",
        "Profile Name: profile7, Profile Type: uereqpdusessrelease",
        "Ue's Passed: 5, Ue's Failed: 0",
        "Profile Status: PASS",
        "Profile Name: profile4, Profile Type: uetriggservicereq",
        "Ue's Passed: 5, Ue's Failed: 0",
        "Profile Status: PASS"

gruyaume · 2024-11-21T12:02:31Z

@gab-arrobo were you able to re-test with the changes in the smf's pfcp adapter changes. I just looked at upfadapter and I can't find anything in there that would break the e2e tests. I believe the problem was in the SMF's pfcp adapter code. In any case, logs and/or captures would be necessary at this point.

I think I did run a test, but it did not work. Let me run another test this afternoon and I can provide logs.

@gruyaume the PRs are now working as expected when using the upfadapter

@thakurajayL, were all of your comments addressed? There is no more comments from my side
        "Profile Name: profile1, Profile Type: register",
        "Ue's Passed: 5, Ue's Failed: 0",
        "Profile Status: PASS",
        "Profile Name: profile2, Profile Type: pdusessest",
        "Ue's Passed: 5, Ue's Failed: 0",
        "Profile Status: PASS",
        "Profile Name: profile3, Profile Type: anrelease",
        "Ue's Passed: 5, Ue's Failed: 0",
        "Profile Status: PASS",
        "Profile Name: profile5, Profile Type: deregister",
        "Ue's Passed: 5, Ue's Failed: 0",
        "Profile Status: PASS",
        "Profile Name: profile7, Profile Type: uereqpdusessrelease",
        "Ue's Passed: 5, Ue's Failed: 0",
        "Profile Status: PASS",
        "Profile Name: profile4, Profile Type: uetriggservicereq",
        "Ue's Passed: 5, Ue's Failed: 0",
        "Profile Status: PASS"

Awesome, the issues were in the smf's pfcp adapter code then/

thakurajayL · 2024-11-26T02:41:22Z

context/datapath.go

@@ -534,37 +476,19 @@ func (dpNode *DataPathNode) ActivateUpLinkPdr(smContext *SMContext, defQER *QER,

 	curULTunnel := dpNode.UpLinkTunnel
 	for name, ULPDR := range curULTunnel.PDR {
-		ULDestUPF := curULTunnel.DestEndPoint.UPF


Why this code is deleted?

thakurajayL · 2024-11-26T02:44:08Z

context/datapath.go

 		ULPDR.QER = append(ULPDR.QER, defQER)

 		// Set Default precedence
 		if ULPDR.Precedence == 0 {
 			ULPDR.Precedence = defPrecedence
 		}

-		var iface *UPFInterfaceInfo


What was motivation to delete this code?

This instance of iface was only used to get the user plane IP address and set the PDR F-TEID accordingly. Now that the F-TEID is set by the UPF, this is not needed anymore

Old code:

var iface *UPFInterfaceInfo if dpNode.IsANUPF() { iface = ULDestUPF.GetInterface(models.UpInterfaceType_N3, smContext.Dnn) } else { iface = ULDestUPF.GetInterface(models.UpInterfaceType_N9, smContext.Dnn) } if upIP, err := iface.IP(smContext.SelectedPDUSessionType); err != nil { logger.CtxLog.Errorf("activate UpLink PDR[%v] failed %v", name, err) return err } else { ULPDR.PDI.SourceInterface = SourceInterface{InterfaceValue: SourceInterfaceAccess} ULPDR.PDI.LocalFTeid = &FTEID{ V4: true, Ipv4Address: upIP, Teid: curULTunnel.TEID, } ULPDR.PDI.UEIPAddress = &ueIpAddr ULPDR.PDI.NetworkInstance = util_3gpp.Dnn(smContext.Dnn) }

thakurajayL · 2024-11-26T02:44:45Z

context/datapath.go

 		}
-
-		if upIP, err := iface.IP(smContext.SelectedPDUSessionType); err != nil {


I think we are removing code which supports iUPF

What's iUPF?

The code you are referring to is not needed anymore as the F-TEID is generated by the UPF.

@gruyaume, FYI, iUPF is the intermediate UPF (N3 and N9 interfaces)

pfcp/adapter/adapter.go

Signed-off-by: Guillaume Belanger <[email protected]>

gruyaume · 2024-11-28T12:52:14Z

@gab-arrobo @thakurajayL can you please review this PR?

feat: request upf to allocate f-teid

78276f1

Signed-off-by: Guillaume Belanger <[email protected]>

gruyaume changed the title ~~feat: request upf to allocate f-teid~~ feat: request UPF to allocate F-TEID Oct 16, 2024

gruyaume added 4 commits October 17, 2024 18:25

Merge branch 'main' into dev-fteid

7432e8a

fix: also implement for downlink pdr

c933860

Signed-off-by: Guillaume Belanger <[email protected]>

fix: set fteid with downlink tunnel

3379f4c

Signed-off-by: Guillaume Belanger <[email protected]>

fix: write f-teid to proper variable

11d07f7

Signed-off-by: Guillaume Belanger <[email protected]>

gruyaume mentioned this pull request Oct 21, 2024

feat: add F-TEID allocation support omec-project/upf#840

Merged

gruyaume added 6 commits November 7, 2024 11:47

Merge branch 'main' into dev-fteid

b0b25e5

chore: add unit tests

f1d3509

Signed-off-by: Guillaume Belanger <[email protected]>

chore: add copyright info

ae36e02

Signed-off-by: Guillaume Belanger <[email protected]>

chore: bump version

19f9281

Signed-off-by: Guillaume Belanger <[email protected]>

Merge branch 'main' into dev-fteid

555120f

chore: use different ports to avoid conflicts

562d5b0

Signed-off-by: Guillaume Belanger <[email protected]>

gruyaume marked this pull request as ready for review November 12, 2024 11:39

gab-arrobo reviewed Nov 12, 2024

View reviewed changes

VERSION Outdated Show resolved Hide resolved

chore: re-add newline to version file

58c4a22

Signed-off-by: Guillaume Belanger <[email protected]>

gruyaume requested a review from gab-arrobo November 12, 2024 15:54

ghislainbourgeois approved these changes Nov 12, 2024

View reviewed changes

dariofaccin approved these changes Nov 13, 2024

View reviewed changes

gab-arrobo reviewed Nov 15, 2024

View reviewed changes

context/datapath_test.go Show resolved Hide resolved

context/upf.go Show resolved Hide resolved

gruyaume added 2 commits November 15, 2024 15:19

feat: store upf generated f-teid when using pfcp adapter

1f33efe

Signed-off-by: Guillaume Belanger <[email protected]>

chore: bump go version to v1.23

f37c030

Signed-off-by: Guillaume Belanger <[email protected]>

gruyaume requested a review from gab-arrobo November 15, 2024 20:20

thakurajayL reviewed Nov 18, 2024

View reviewed changes

context/datapath.go Show resolved Hide resolved

thakurajayL reviewed Nov 18, 2024

View reviewed changes

context/datapath.go Show resolved Hide resolved

pfcp/adapter/adapter.go Show resolved Hide resolved

pfcp/adapter/adapter.go Outdated Show resolved Hide resolved

pfcp/handler/handler.go Show resolved Hide resolved

Merge branch 'main' into dev-fteid

51db900

chore: add missing lock statement

ff1ded0

Signed-off-by: Guillaume Belanger <[email protected]>

gruyaume requested a review from thakurajayL November 18, 2024 13:51

gab-arrobo reviewed Nov 19, 2024

View reviewed changes

go.mod Outdated Show resolved Hide resolved

Merge branch 'main' into dev-fteid

c964b02

gruyaume requested a review from gab-arrobo November 19, 2024 14:09

Merge branch 'main' into dev-fteid

204c912

thakurajayL reviewed Nov 26, 2024

View reviewed changes

chore: defer mutex unlock right away

912b4e5

Signed-off-by: Guillaume Belanger <[email protected]>

gruyaume requested a review from thakurajayL November 26, 2024 17:39

thakurajayL approved these changes Nov 28, 2024

View reviewed changes

thakurajayL merged commit ec30f3e into omec-project:main Nov 28, 2024
8 checks passed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

feat: request UPF to allocate F-TEID #334

feat: request UPF to allocate F-TEID #334

gruyaume commented Oct 16, 2024 •

edited

Loading

thakurajayL commented Oct 19, 2024

gab-arrobo left a comment

gruyaume commented Nov 15, 2024

gab-arrobo commented Nov 16, 2024

gruyaume commented Nov 18, 2024

gruyaume commented Nov 18, 2024

gruyaume commented Nov 20, 2024

gab-arrobo commented Nov 20, 2024

gab-arrobo commented Nov 21, 2024

gruyaume commented Nov 21, 2024

thakurajayL Nov 26, 2024

thakurajayL Nov 26, 2024

gruyaume Nov 26, 2024 •

edited

Loading

thakurajayL Nov 26, 2024

gruyaume Nov 26, 2024 •

edited

Loading

gab-arrobo Nov 27, 2024

gruyaume commented Nov 28, 2024

		}

		if upIP, err := iface.IP(smContext.SelectedPDUSessionType); err != nil {

feat: request UPF to allocate F-TEID #334

feat: request UPF to allocate F-TEID #334

Conversation

gruyaume commented Oct 16, 2024 • edited Loading

Description

Context

Reference

thakurajayL commented Oct 19, 2024

gab-arrobo left a comment

Choose a reason for hiding this comment

gruyaume commented Nov 15, 2024

gab-arrobo commented Nov 16, 2024

gruyaume commented Nov 18, 2024

gruyaume commented Nov 18, 2024

gruyaume commented Nov 20, 2024

gab-arrobo commented Nov 20, 2024

gab-arrobo commented Nov 21, 2024

gruyaume commented Nov 21, 2024

thakurajayL Nov 26, 2024

Choose a reason for hiding this comment

thakurajayL Nov 26, 2024

Choose a reason for hiding this comment

gruyaume Nov 26, 2024 • edited Loading

Choose a reason for hiding this comment

thakurajayL Nov 26, 2024

Choose a reason for hiding this comment

gruyaume Nov 26, 2024 • edited Loading

Choose a reason for hiding this comment

gab-arrobo Nov 27, 2024

Choose a reason for hiding this comment

gruyaume commented Nov 28, 2024

gruyaume commented Oct 16, 2024 •

edited

Loading

gruyaume Nov 26, 2024 •

edited

Loading

gruyaume Nov 26, 2024 •

edited

Loading