Skip to content

Commit

Permalink
[ycabled] add notification for gRPC connection state transitions to I…
Browse files Browse the repository at this point in the history
…DLE/TRANSIENT_FAILURE (#295)

Signed-off-by: vaibhav-dahiya vdahiya@microsoft.com
For the cases where gRPC connectivity to server becomes IDLE/TRANSIENT_FAILURE, the ToR should periodically query/retry establish admin state and revert the Forwarding state back to active-active for both the T0's
This PR tries to attain that via adding a transient failure message to APP DB when connectivity is lost, so each time connectivity state changes back to not desired state(IDLE/TRANSIENT_FAILURE) we log a message, and linkmgr will query again to get gRPC in sync

Description
Motivation and Context
How Has This Been Tested?
Unit-tests and deploying changes to testbed

Additional Information (Optional)
  • Loading branch information
vdahiya12 authored Sep 21, 2022
1 parent 4b3b238 commit 6522c46
Showing 1 changed file with 21 additions and 1 deletion.
22 changes: 21 additions & 1 deletion sonic-ycabled/ycable/ycable_utilities/y_cable_helper.py
Original file line number Diff line number Diff line change
Expand Up @@ -424,16 +424,36 @@ def connect_channel(channel, stub, port):
def create_channel(type, level, kvp, soc_ip, port):


#Helper callback to get an channel connectivity state
appl_db = {}
fwd_state_response_tbl = {}
namespaces = multi_asic.get_front_end_namespaces()
for namespace in namespaces:
# Open a handle to the Application database, in all namespaces
asic_id = multi_asic.get_asic_index_from_namespace(namespace)
appl_db[asic_id] = daemon_base.db_connect("APPL_DB", namespace)
fwd_state_response_tbl[asic_id] = swsscommon.Table(
appl_db[asic_id], "FORWARDING_STATE_RESPONSE")

asic_index = y_cable_platform_sfputil.get_asic_id_for_logical_port(port)

# Helper callback to get an channel connectivity state
def wait_for_state_change(channel_connectivity):
if channel_connectivity == grpc.ChannelConnectivity.TRANSIENT_FAILURE:
helper_logger.log_notice("gRPC port {} state changed to TRANSIENT_FAILURE".format(port))
# for connectivity state to FAILURE/IDLE report a failure
fvs_updated = swsscommon.FieldValuePairs([('response', 'failure')])
fwd_state_response_tbl[asic_index].set(port, fvs_updated)

if channel_connectivity == grpc.ChannelConnectivity.CONNECTING:
helper_logger.log_notice("gRPC port {} state changed to CONNECTING".format(port))
if channel_connectivity == grpc.ChannelConnectivity.READY:
helper_logger.log_notice("gRPC port {} state changed to READY".format(port))
if channel_connectivity == grpc.ChannelConnectivity.IDLE:
helper_logger.log_notice("gRPC port {} state changed to IDLE".format(port))
# for connectivity state to FAILURE/IDLE report a failure
fvs_updated = swsscommon.FieldValuePairs([('response', 'failure')])
fwd_state_response_tbl[asic_index].set(port, fvs_updated)

if channel_connectivity == grpc.ChannelConnectivity.SHUTDOWN:
helper_logger.log_notice("gRPC port {} state changed to SHUTDOWN".format(port))

Expand Down

0 comments on commit 6522c46

Please sign in to comment.