Whats the best version to have on p10 bh’s?
I have a bh20 8.2.2 on both ends, it link tests at 10.6 down 3.3 up and I monitor the average traffic well below 3/1. during peak hours this bh latency goes from 7 to 60ms and speeds drop dramatically. I monitor the bandwidth usage at around 6/2 during peak hours.
Anyone else know why the performance varies so much while under load?
How are you monitoring the latency?
8.x software does not respond well to pings under load or otherwise. Do not rely on that to give a good indication. Ping something else. I run 8.2.2 on my BH links and they seem to be doing just fine. I however do little snmp polling of any 8.x device at this time since it tends to cause too many problems.
well when customers call complaining about speed issues I ping the master,slave and an AP.The high latency starts at the bh slave, which usually indicates a bottle neck. the latency to the AP’s behind the slave are all about the same, 2-5ms higher then to the slave.
When the link is not under load it runs at around 7ms - 0 percent loss
when I get calls the link runs around 40ms - 4 percent loss
This link was established with a p9 link prior to the fires in southern CA. I was forced to replace the link with p10 boards on firmware 8.*, this link has never been as good as it was before the fire.
vince wrote: well when customers call complaining about speed issues I ping the master,slave and an AP.The high latency starts at the bh slave, which usually indicates a bottle neck. the latency to the AP's behind the slave are all about the same, 2-5ms higher then to the slave.
When the link is not under load it runs at around 7ms - 0 percent loss
when I get calls the link runs around 40ms - 4 percent loss
This link was established with a p9 link prior to the fires in southern CA. I was forced to replace the link with p10 boards on firmware 8.*, this link has never been as good as it was before the fire.
YES, I got the same result too, after upgrade all our P9 BH to 8.2.2
The ping is very not stable, compare to 7.36 even twice the bh is locked up, and we are crazy to find out what happened.
I asked the technician to climb up to default the bh and before the technician reached the bh, I checked all the ethernet speed at configuration page (at slave) and it suddenly worked :x
What I did than, change all the BH to 736 and all the funny funny things gone....gone forever.....
I never touch all of my BH again..... i hate you motorola :evil:
Btw, did they ever look at this forum? supposed they hire 1 tier1 support to handle this forum. Maybe someone from US can push motorola to put 1 tier1 support at this forum so the higher tier support engineer know all the issues, bugs and recommendation from the people at this forum who pay their salary.... :evil:
… . moto states clearly in several documents that ping packets are given low priority during operation while under high load.
if you want an accurate number, ping through it to a device on the other side.
otherwise there are several tools you can use to monitor latency/stability which are far superior to ping.
keep in mind, ping is the most basic link test. if you’re mapping your network and benchmarking performance based on pings, you have some homework to do.
there’s many aspects of network performance to track, such as broadcast/multicast activity, packet fragmentation (mtus!!!), and packet switching rates vs data rates to name a few…
pings can tell you little to nothing about any of those.