Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[CRM][MPLS NHOP] Fix the mpls nexthop CRM attribute #2008

Merged
merged 1 commit into from
Nov 5, 2021

Conversation

smaheshm
Copy link
Contributor

@smaheshm smaheshm commented Nov 4, 2021

What I did
Fix the mpls nexthop CRM attribute

MPLS v4 nexthops will share the same resource as IPv4 NHOP, there is no separate resource partition for MPLS nexthop.

Why I did it
Getting the following error in logs:
Oct 29 22:40:03.679316 str2-7804-lc3-1 ERR swss#orchagent: :- getResAvailableCounters: Failed to get availability for object_type 63 , rv:-5
.
.

How I verified it
Verified the message is not produced anymore

Details if related

@smaheshm smaheshm requested review from kcudnik and prsunny November 4, 2021 21:27
@@ -64,7 +64,7 @@ const map<CrmResourceType, uint32_t> crmResSaiAvailAttrMap =
{ CrmResourceType::CRM_SNAT_ENTRY, SAI_SWITCH_ATTR_AVAILABLE_SNAT_ENTRY },
{ CrmResourceType::CRM_DNAT_ENTRY, SAI_SWITCH_ATTR_AVAILABLE_DNAT_ENTRY },
{ CrmResourceType::CRM_MPLS_INSEG, SAI_OBJECT_TYPE_INSEG_ENTRY },
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Is the INSEG_ENTRY correct?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

yes, label routes usually have a separate tables than IP routes.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

but in code it doesn't look right, will check further.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

There is no SAI attribute for AVAILABLE INSEG entries. How are such cases handled. Should the code for inseg label CRM be removed?

@smaheshm smaheshm merged commit 01c243a into sonic-net:master Nov 5, 2021
EdenGri pushed a commit to EdenGri/sonic-swss that referenced this pull request Feb 28, 2022
#### What I did

Fixes sonic-net#2007 

Most of the children of `/BGP_NEIGHBOR/*` except `admin_status` are create-only field i.e. they can only be created with the neighbor but cannot be modified later.

Validated each attribute is read-only by the following steps:
* Delete a neighbor
* Add the neighbor back without the attribute under test e.g. `holdtime`
* show running config for the neighbor
* show neighbor config using `show ip bgp neighbor <ip>`
* Add just the attribute under test e.g. `holdtime`
* show running config for the neighbor -- we can see the attribute is added
* show neighbor config using `show ip bgp neighbor <ip>` -- we can see the attribute change did not take effect

Example for `holdtime`:
```sh
admin@vlab-01:~$ sudo config apply-patch remove-bgp-neighbor.json -i '' 
.
.
.
Patch applied successfully.
admin@vlab-01:~$ sudo config apply-patch remove-bgp-neighbor.json -i ''
.
.
.
Error: can't remove a non-existent object '10.0.0.57'
admin@vlab-01:~$ sudo config apply-patch add-bgp-neighbor-without-holdtime.json -i ''
Patch Applier: Patch application starting.
Patch Applier: Patch: [{"op": "add", "path": "/BGP_NEIGHBOR/10.0.0.57", "value": {"admin_status": "up", "asn": "64600", "keepalive": "3", "local_addr": "10.0.0.56", "name": "ARISTA01T1", "nhopself": "0", "rrclient": "0"}}]
.
.
.
Patch applied successfully.
admin@vlab-01:~$ show runningconfiguration all | grep 10.0.0.57 -A8
        "10.0.0.57": {
            "admin_status": "up",
            "asn": "64600",
            "keepalive": "3",
            "local_addr": "10.0.0.56",
            "name": "ARISTA01T1",
            "nhopself": "0",
            "rrclient": "0"
        },
admin@vlab-01:~$ show ip bgp neighbors 10.0.0.57
.
.
. 
  Hold time is 180, keepalive interval is 3 seconds
.
.
. 
admin@vlab-01:~$ sudo config apply-patch add-holdtime.json -i ''
Patch Applier: Patch application starting.
Patch Applier: Patch: [{"op": "add", "path": "/BGP_NEIGHBOR/10.0.0.57/holdtime", "value": "10"}]
.
.
. 
Patch applied successfully.
admin@vlab-01:~$ show runningconfiguration all | grep 10.0.0.57 -A10
        "10.0.0.57": {
            "admin_status": "up",
            "asn": "64600",
            "holdtime": "10",
            "keepalive": "3",
            "local_addr": "10.0.0.56",
            "name": "ARISTA01T1",
            "nhopself": "0",
            "rrclient": "0"
        },
        "10.0.0.59": {
admin@vlab-01:~$ show ip bgp neighbors 10.0.0.57
BGP neighbor is 10.0.0.57, remote AS 64600, local AS 65100, external link
.
.
. 
  Hold time is 180, keepalive interval is 3 seconds
.
.
. 
admin@vlab-01:~$ 
```

Also added a validation to `create-only` fields to reject moves that add their parents without them, because we would have to delete their parents again later and add it back. There is no point.
Example assume we have 2 fields marked with create-only namely x,y and they are under c. 
The patch would be:
```
{"op":"add", "path":"/a/b/c", "value":{"x":"value_x", "y":"value_y"}}
```
The generated moves would be:
```
{"op":"add", "path":"/a/b/c", "value":{"x":"value_x"}}
{"op":"remove", "path":"/a/b/c"}
{"op":"add", "path":"/a/b/c", "value":{"x":"value_x", "y":"value_y"}}
```

There is no point of the first 2 moves, because the `y` is create only and it will require the object to be deleted again then added. 


#### How I did it
Marked the fields as create only

#### How to verify it
unit-test

#### Previous command output (if the output of a command-line utility has changed)

#### New command output (if the output of a command-line utility has changed)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants