Guys, we have our main backhaul master (from First hop tower to NOC) showing the following stats:
inoctects Count : 390270516
inucastpkts Count : 24390931
Innucastpkts Count : 125049
indiscards Count : 0
inerrors Count : 0
inunknownprotos Count : 0
outoctets Count : 783437671
outucastpktsCount : 22588542
outnucastpkts Count : 196810
outdiscards Count : 324014
outerrors Count : 0
The outdiscards count is bothering me.
What do you think is causing this? An oversaturation on the Tower > NOC link?
The connection is currently set at 35% downlink.
On this same BH pair we upgraded to 8.2 this night. Since the upgrade we have had three outages with very strange symptoms.
The BHS remains registered but we lose ping to the BHM and so the link is effectively down until we reboot the BHS.
Nothing in the logs that could give us any indication of what is happening.
8.2 is NOT recommended for backhauls yet. If you read the release notes Moto states that there will be a point release for 8.2 to be used on backhauls. if the backhaul is a P10 and shipped with version 8 you should leave it at the version it shipped with. if it is a P9 or P8 then 7.3.6 is recommended (some people on the forums even suggest 7.1.4 for backhauls. we have all of ours at 7.3.6 with no issues)
Hi.
Moto has released a second version of 8.2 which is now suitable for Backhauls (as per the release notes).
The first version did NOT allow you to install onto BH units. The second one does.
I would try setting the ethernet link to 10Mbps full and half and uncheck 100Mbps full and half. sometimes a bad cable or crimp will give you errors and slowing down the link can compensate for that. the BH units can’t push data faster than 10Mbps over the air interface so you won’t be loosing any functionality.
The cables are fine… no ping loss.
The BH20 can push up to 14 megabits so limiting it to 10 megabit is a waste of 35% of the bandwidth.
thanks for the feedback.
The 14Mbps is aggregate throughput which if you have the downlink set to say 75% then that means you get 10Mbps BHM -> BHS and 4Mbps BHS -> BHM. if you set the downlink to 50% then you get 7Mbps up and down.
i would check the cables with a cable tester and/or set the ethernet speed slower and see if that gets rid of the errors.
Yes, you’re right about the speed if you set the ratios beneath 10 megabits in any direction.
Checked the cabes, they are fine. Aready did the downgrade to 10 megabit on the Ethernet with no luck.
Are you graphing this link? What is the sustained traffic?
Try this procedure:
At 8am reboot the radios. At 8pm note the ratio of outdiscards to outoctets. Reboot again. At 8 am note the ratio again.
If the ratio is substantially higher during the day it’s probably saturated.
Have also seen mismatches with auto negotiate switches and Canopy. Try forcing the switch port and the radio to 100M/full.
bypass any cat 5 suppressors and test again.
I replaced this link with a PTP600 running at 150 Megabit…
Solved the issue… so i guess the link was being maxed out.