Support migration from the Firefly LNS #97
Conversation
KrishnaIyer
requested review from
happyRip,
johanstokking and
adriansmares
as code owners
There was a problem hiding this comment.
DEBUG Failed to determine channel index of uplink {"ack": false, "adr": false, "adr_ack_req": false, "band_id": "EU_863_870", "bandwidth": 125000, "class_b": false, "data_rate": "lora:{bandwidth:125000 spreading_factor:7 coding_rate:"4/5"}", "dev_addr": "29559735", "error": "error:pkg/networkserver:uplink_channel_not_found (uplink channel not found)", "f_cnt_reset": false, "f_opts_len": 0, "f_port": 1, "frequency": 867700000, "full_f_cnt_up": 11, "grpc.method": "HandleUplink", "grpc.service": "ttn.lorawan.v3.GsNs", "last_f_cnt_up": 10, "m_type": "UNCONFIRMED_UP", "mac_count": 0, "mac_version": "MAC_V1_0_2", "major": "LORAWAN_R1", "namespace": "networkserver", "pending_session": false, "phy_payload_len": 14, "received_at": 1698777742.7678819, "request_id": "01HE3GPFDFN4CBZX9Q697JE83A", "spreading_factor": 7, "uplink_f_cnt": 11}
This is normal to a certain extent - you don't pass any MAC state while importing the end device, which means that the LNS builds a MAC state which has the current parameters with only the 'boot' channels - 867700000 is not a 'boot' channel.
The ttnv2 source can show you how to build the MAC state and how to edit it. In this case you can do the same current = desired trick and then work back the different parameters in Firefly.
There was a problem hiding this comment.
This JoinEUI thing is probably quite problematic, no? This only works if there's one JoinEUI per batch? And you either migrate one device, or you migrate all devices accessible with the API key?
I presume Firefly just takes the JoinEUI from the join-request, and uses the DevEUI as primary identifier?
So if there's a mix of JoinEUIs in a larger deployment, devices need to be migrated one by one?
| logLevel := log.InfoLevel | ||
| if verbose, _ := cmd.Flags().GetBool("verbose"); verbose { | ||
| logLevel = log.DebugLevel | ||
| } | ||
| logHandler, err := log.NewZap("console") | ||
| if err != nil { | ||
| return err | ||
| } | ||
| logger = log.NewLogger( | ||
| logHandler, | ||
| log.WithLevel(logLevel), | ||
| ) | ||
| rpclog.ReplaceGrpcLogger(logger) | ||
| ctx = log.NewContext(ctx, logger) | ||
|
|
||
| exportCfg.DevIDPrefix, _ = cmd.Flags().GetString("dev-id-prefix") | ||
| exportCfg.EUIForID, _ = cmd.Flags().GetBool("set-eui-as-id") | ||
| ctx = export.NewContext(ctx, exportCfg) | ||
|
|
||
| cmd.SetContext(ctx) |
There was a problem hiding this comment.
The context was previously never propagated to the subcommands and hence this logger and the exportconfig were never used.
I've fixed the export config but we don't need this logger since we use Zap in the source.
|
Retested the changes
|
Yes indeed. The way I have documented this is to migrate devices in batches by creating a file with the Device EUIs and piping that into the |
Summary
Support migration from the Firefly LNS. Replaces #49.
Changes
Testing
With a The Things Uno.
This can also be checked via the UI
Regressions
NA. This is new functionality.
Notes for Reviewers
EU_863_870_TTNinstead? I'm using a The Things Uno for testing.Does anything stand out to you here @adriansmares?
Checklist
CHANGELOG.md.