@Jeff-UK so I discovered today loss in another situation…
Device and App registered to EU1v3, one gateway in reach, also registered to EU1v3.
Gateway is a Mikrotik UDP gateway again (yes I know, UDP…) with LTE (yes I know, LTE, baaaad!).
I see the uplink from the node in the TTN gateway ‘Live data’ tab, but not in the TTN device ‘Live tab’.
Also, the packet is missing on all my three Webhook endpoints (which are in different datacenters).
So for me it looks cleary that TTN is loosing messages within their own system which should not happen, at least not on a regular basis… and no packet broker involved… and yes, the packet arrived at TTN side 
Metadata of the correctly processed uplink:
{
"name": "gs.up.receive",
"time": "2021-08-31T20:03:49.745908419Z",
"identifiers": [
{
"gateway_ids": {
"gateway_id": "hir-ttn01v3"
}
},
{
"gateway_ids": {
"gateway_id": "hir-ttn01v3",
"eui": "4836372047001D00"
}
}
],
"data": {
"@type": "type.googleapis.com/ttn.lorawan.v3.UplinkMessage",
"raw_payload": "QEIKCyYA3CcBE01ET2RMCTrWtOHygexSMP0=",
"payload": {
"m_hdr": {
"m_type": "UNCONFIRMED_UP"
},
"mic": "7FIw/Q==",
"mac_payload": {
"f_hdr": {
"dev_addr": "260B0A42",
"f_ctrl": {},
"f_cnt": 10204
},
"f_port": 1,
"frm_payload": "E01ET2RMCTrWtOHygQ=="
}
},
"settings": {
"data_rate": {
"lora": {
"bandwidth": 125000,
"spreading_factor": 7
}
},
"coding_rate": "4/5",
"frequency": "868300000",
"timestamp": 2394773225,
"time": "2021-08-31T20:03:50.991572Z"
},
"rx_metadata": [
{
"gateway_ids": {
"gateway_id": "hir-ttn01v3",
"eui": "4836372047001D00"
},
"time": "2021-08-31T20:03:50.991572Z",
"timestamp": 2394773225,
"rssi": -73,
"channel_rssi": -73,
"snr": 9.25,
"uplink_token": "ChkKFwoLaGlyLXR0bjAxdjMSCEg2NyBHAB0AEOmt9fUIGgwIpZa6iQYQ8dDP4wIgqPztnNnrHQ==",
"channel_index": 1
}
],
"received_at": "2021-08-31T20:03:49.745793649Z",
"correlation_ids": [
"gs:conn:01FEB3H9WXPK7GD0AZ7K6AZVRK",
"gs:uplink:01FEEWZ2VHYCPYDT6SE3J9PJYT"
]
},
"correlation_ids": [
"gs:conn:01FEB3H9WXPK7GD0AZ7K6AZVRK",
"gs:uplink:01FEEWZ2VHYCPYDT6SE3J9PJYT"
],
"origin": "ip-10-100-5-46.eu-west-1.compute.internal",
"context": {
"tenant-id": "CgN0dG4="
},
"visibility": {
"rights": [
"RIGHT_GATEWAY_TRAFFIC_READ",
"RIGHT_GATEWAY_TRAFFIC_READ"
]
},
"unique_id": "01FEEWZ2VHFM930T1MTXMT9QFX"
}
Metadata of the missing uplink:
{
"name": "gs.up.receive",
"time": "2021-08-31T20:08:49.148113249Z",
"identifiers": [
{
"gateway_ids": {
"gateway_id": "hir-ttn01v3"
}
},
{
"gateway_ids": {
"gateway_id": "hir-ttn01v3",
"eui": "4836372047001D00"
}
}
],
"data": {
"@type": "type.googleapis.com/ttn.lorawan.v3.UplinkMessage",
"raw_payload": "QEIKCyYA3ScBkpJWAk1cyoAv+Gti376IhpE=",
"payload": {
"m_hdr": {
"m_type": "UNCONFIRMED_UP"
},
"mic": "voiGkQ==",
"mac_payload": {
"f_hdr": {
"dev_addr": "260B0A42",
"f_ctrl": {},
"f_cnt": 10205
},
"f_port": 1,
"frm_payload": "kpJWAk1cyoAv+Gti3w=="
}
},
"settings": {
"data_rate": {
"lora": {
"bandwidth": 125000,
"spreading_factor": 7
}
},
"coding_rate": "4/5",
"frequency": "868500000",
"timestamp": 2694196729,
"time": "2021-08-31T20:08:50.415727Z"
},
"rx_metadata": [
{
"gateway_ids": {
"gateway_id": "hir-ttn01v3",
"eui": "4836372047001D00"
},
"time": "2021-08-31T20:08:50.415727Z",
"timestamp": 2694196729,
"rssi": -78,
"channel_rssi": -78,
"snr": 8.25,
"uplink_token": "ChkKFwoLaGlyLXR0bjAxdjMSCEg2NyBHAB0AEPnb2IQKGgsI0Zi6iQYQrZnIRiCoqY7VtPQd",
"channel_index": 2
}
],
"received_at": "2021-08-31T20:08:49.147983533Z",
"correlation_ids": [
"gs:conn:01FEB3H9WXPK7GD0AZ7K6AZVRK",
"gs:uplink:01FEEX877WCQEYNSYM67DA29CP"
]
},
"correlation_ids": [
"gs:conn:01FEB3H9WXPK7GD0AZ7K6AZVRK",
"gs:uplink:01FEEX877WCQEYNSYM67DA29CP"
],
"origin": "ip-10-100-5-46.eu-west-1.compute.internal",
"context": {
"tenant-id": "CgN0dG4="
},
"visibility": {
"rights": [
"RIGHT_GATEWAY_TRAFFIC_READ",
"RIGHT_GATEWAY_TRAFFIC_READ"
]
},
"unique_id": "01FEEX877W5FJJ11Z7ED57Z1CS"
}
And the screenshots:
Gateway live data:
Device live data:
I believe your Laird can support BS (though do not know your firmware revision so may need an update!), it may be useful to try that to see if matters improve for you. If your nodes are on V3 then your V2 GW’s will be servicing them by delivering packets through the PacketBroker, using LTE is generally fine for backhaul if in a decent reception area (I use both 3G & 4G connections in part of my GW fleet, with little problem), but I know of some folk who have struggled with additional latency through cellular connections. In the case of the Mikrotik that latency will be added to the additional latency through PacketBroker so would be really good to get more details and logs. When you do see data arrive in your applications (or on the device traffic tab) in the Console does the metadata show that all three GW’s are servicing each message or is it patchy with perhaps one GW catching most with the other two missing the odd message or appearing to not contribute (I suspect it is poss that if one GW’s handles quickly, and the other two are on long latency/delay then NS may ignore late arriving messages as part of its de-dupe processes and window (I do not know the window set up).
They came back to me 2 months later and apologised when whole shelf went down! (Known the guys for years… this was second major problem I called out…they listen now! 
)
is that even possible), has node been migrated into V3 but also still active in V2…
