Repeated ignition attempts

Jason_Cox · March 18, 2026, 8:46pm

Does it indicate a problem if the number of ignition attempts on our links is climbing? If so, what should we look for to correct it?

We have a V5000 DN that has two V1000 CNs connected to it. One of the CNs shows a link time of 1h 38m and 854 ignition attempts. The other shows a link time of 57m and 56 ignition attempts. These are all on software version 1.7. The signal strengths are below. The one with the lowest SNR is actually not the one with the most ignition attempts.

Name	Direction	MCS	RSSI	SNR
link-PFALLS-MAINT-DN-PFALLS-PPO	PFALLS-MAINT-DN to PFALLS-PPO	9	-68 dBm	5 dB
link-PFALLS-INTERP-PFALLS-MAINT-DN	PFALLS-INTERP to PFALLS-MAINT-DN	9	-60 dBm	14 dB
link-PFALLS-INTERP-PFALLS-MAINT-DN	PFALLS-MAINT-DN to PFALLS-INTERP	8	-61 dBm	12 dB
link-PFALLS-MAINT-DN-PFALLS-PPO	PFALLS-PPO to PFALLS-MAINT-DN	2	-61 dBm	13 dB

Antony · March 19, 2026, 10:45am

Not quite enough data to see what’s going but what do we see…

The DN setup, are you using one sector or two: If two are they both the same polarity and what frequency are they on? Just eliminate any potential issues by keeping the polarity the same but pick different channels if they are on different sectors. I assume you have a good GPS signal.

Looking at the links: to really understand what’s going on I would look at the Tx power and the beam angles, make sure you’re on the latest software version 1., plus look at the graphs, ideally through cnMaestro to see how things change over time.

Your two link, we can see:

PFALLS-MAINT-DN to PFALLS-INTERP is pretty well balanced, RSSI running at -60/-61, however given the MCS in this snapshot is 8, tso may need a closer look but you still have 12dB margin so not too worried.

However,

PFALLS-MAINT-DN to PFALLS-PPO, is unbalanced, you have MCS 9 in one direction and MCS 2 in the other.

Now the low MCS, I would look at the receiving end, is there anything in the Fresnel that could be causing reflections or obstructions? Check the beam angle, does the Tx and RX match, do they change over time (Note in cnMaestro 1.6.1 we’re adding a graph of the angle) until then you can only see the current angles.

Looking in the other direction, whilst you at -68dBm and 5 dB SNR, what’s the power and range?

Another reason for an out of balance Tx/Rx is interference, hence check the DN config if there’s no other device or reflections.

All this information can be gather from the cnMaestro graphs. Raise a ticket, send us debug. Seems like an interesting one for showing people what to look for…

Jason_Cox · March 19, 2026, 3:39pm

The V5000 DN has two sectors. Auto channel selection had placed them on the same channel. I manually set one sector to channel 1 and one to channel 2 to see what happens.So far the links are stable, and haven’t dropped in over an hour.

The distances of the links are: Maint→ Interp: about 36 meters. Maint→ PPO: about 45 meters. For some reason, Topology is now reporting the distance at over 300 meters. I made sure the accuracy was set to 10,000 meters for the devices. The DN keeps changing itself to around 5.7 meters accuracy. Does the DN determine its accuracy dynamically? It doesn’t seem to be causing a problem.

GPS shows 20 satellites tracked.

I downloaded the link statistics, and the TX and RX angles match. I downloaded them several minutes apart, and the angles didn’t change.

The LOS isn’t perfect. There are a couple of small trees that cause partial fresnel obstructions.

Right now the TX and RX SNR and MCS rates are closer together.

link-PFALLS-MAINT-DN-PFALLS-PPO	PFALLS-MAINT-DN to PFALLS-PPO	9	-61 dBm	12 dB
link-PFALLS-MAINT-DN-PFALLS-PPO	PFALLS-PPO to PFALLS-MAINT-DN	9	-61 dBm	12 dB
link-PFALLS-INTERP-PFALLS-MAINT-DN	PFALLS-INTERP to PFALLS-MAINT-DN	2	-70 dBm	2 dB
link-PFALLS-INTERP-PFALLS-MAINT-DN	PFALLS-MAINT-DN to PFALLS-INTERP	2	-68 dBm	4 dB

Antony · March 19, 2026, 5:04pm

Thanks Jason,

The v5000 has GPS in it so that will position itself accurately and override any setting you’ve put in as the accuracy it gets from the satellites is going to be better than 10000m. This is why we choose this as a default to ensure it’s overridden. You can force a location by having this at 0.

The V1000’s however don’t have GPS so you need to position those. We do use the position for channel and golay optimisation but on big networks as you roll out and what to let the E2E plan interference mitigation.

You can use the same channel, if the polarisation is the same, I assume that’s the case, unless the links are close to the intersection of the two sectors, then space the channels out. I would use channel 2 and 4 if this is the only DN and in PMP mode (hub and spoke).

The ranges you are using MCS 2 at 36m, you have a bigger LOS issue than you think, there’s some wrong. Raise a ticket and we’ll get someone to look at your data.

Do you have a picture, especially from the DN side point to the CN INTERP.

Prasanna_TM · March 20, 2026, 3:28pm

Hi Jason,

Just to add to Antony’s observations:

The higher ignition attempt count you’re seeing is typically associated with repeated link establishment or beamforming retries, and in your case, it appears to correlate with the weaker link:

The INTERP link is operating at MCS 2 with ~2–4 dB SNR, which is quite low for 60 GHz
At this SNR level, the link may remain up, but can experience frequent BF retraining / ignition retries, contributing to the higher ignition count

Also, your earlier configuration with both sectors on the same channel could have introduced interference, which may have further increased ignition attempts. Your change to separate channels is the right step.

What you can check further:

Monitor whether ignition attempts continue to increase after channel separation
Observe BF training stats (bfTrainingReq / bfTrainingRsp)
Node GUI –> Diagnostics –> Engineering logs –> e2e_minion logs
Try improving LOS for the INTERP link (even minor Fresnel clearance improvements can significantly help at 60 GHz)

Key takeaway:

The ignition attempts in this case are likely symptomatic of marginal RF conditions (low SNR + partial obstruction) rather than a standalone issue.

Jason_Cox · March 20, 2026, 8:28pm

Prasanna, putting the sectors on separate channels seems to have helped. There hasn’t been any more ignition attempts in over 24 hours.

I tried different combinations of channels, and it appears to be working best with sector 1 on channel 4 and sector 2 on channel 2. The SNRs are now better on the Interp. link at 6 dB and 9 dB. The MCS is still lower at 4 and 2. We might be able to ask someone at the site to trim trees if we have any more trouble.