Still trying to track this latency issue but need to gauge my expectations…
Moto claim a latency of 2.5ms between backhauls… my set up
Cisco1 3500=>(BH1)----(BH2)=>CMM1<=(BH3)----(BH4)=>CMM2<=(BH5)—(BH6)=>Cisco2 3500
Tests results based on 1000 pings, all tests done from Cisco 1, results format (min/avg/max), the gateway for all the kit sits on Cisco1
ping BH1 ==== 1/5/144
ping BH2 ==== 1/12/236
ping cmm1 === 1/11/92
ping BH3 ==== 4/14/132
ping BH4 ==== 4/21/144
ping cmm2 === 4/21/140
ping BH5 ==== 8/26/196
ping BH6 ==== 8/32/216
ping cisco2 === 8/32/140
Does this sound ok ? Obviously I need to repeat the tests to build a real picture. Also these results have been done at a very quite time (sat morning).
What you guys think ?
144 ms from your Cisco interface to the Ethernet interface of BH1 is pretty high. Is the BH wired directly to a FE or E interface of the 3500, or is there a switch in between? Regardless, you should be seeing nothing more than ~10 ms there during a very quiet time.
Login to Cisco 3500-1 and switch to enable mode, then issue a “sh interfaces” command and make sure you have zero to minimal CRC errors, etc from your first Cisco to your first BH.
If all looks OK there, then look into RF. I still think 144 ms is awfully high.
I agree…
sh int…
207507 input errors, 207507 CRC, 0 frame, 0 overrun, 0 ignored
damn CRC errors… damn cables… going to clear counters and see what happens…
does the RF look ok ? to me doesn’t fit the moto spec but then again nothing ever does…
Which BH’s are masters and which are slaves? Do the CMM’s seem to be functioning properly?
I would definitely check that cabling from BH1 to your Cisco 3500-1. If you are seeing that much latency on the cable, its going to add up when traversing the entire path. I would also check all the other cabling along the way, look at the Ethernet Stats on all the BH’s.
The other thing you can do to check the links and only the links is to physically visit each BHM, telnet to the radio, and from the telnet interface, ping the remote BHS. Obviously try to do it during a maintenance window.
Keep me updated. I have had issues with Best Tronics Outdoor Shielded Cable on various installations as well as bad Ethernet Ports on BHs.
we normally make our own cable… I will be going through a clean up and checking all the cables…
When you say telnet into the BHM and telenet, how will that test the link, you only get a.c.d.e is alive
BH2 is a master / BH1 is a slave
The BH1 is not reporting any CRC
BH1 has bee nup 38 days
inoctects Count 1637309667
inucastpkts Count 2692578496
Innucastpkts Count 206063241
indiscards Count 0
inerrors Count 6627
inunknownprotos Count 0
outoctets Count 2094908805
outucastpktsCount 2363464740
outnucastpkts Count 145705404
outdiscards Count 4937
outerrors Count 154
RxBabErr 0
TxHbErr 0
EthBusErr 0
CRCError 0
RxOverrun 6627
LateCollision 0
RetransLimitExp 0
TxUnderrun 154
CarSenseLost 0
BH2 has been up 25 days
inoctects Count 1905178989
inucastpkts Count 1565129803
Innucastpkts Count 101998597
indiscards Count 0
inerrors Count 1982
inunknownprotos Count 0
outoctets Count 11138260
outucastpktsCount 1775743462
outnucastpkts Count 138519342
outdiscards Count 135809
outerrors Count 140
RxBabErr 0
TxHbErr 0
EthBusErr 0
CRCError 0
RxOverrun 1982
LateCollision 0
RetransLimitExp 0
TxUnderrun 140
CarSenseLost 0
You are right, I forgot the CLI’s of the radios do not report times when you ping. Sorry about that.
So your Cisco interface shows errors but the BH’s Ethernet Interface does not? Sounds identical to a problem I had.
A five-legged BH network came to a central point where all the BHM’s and the CMM Micro was. One of the BHM’s was reporting a lot of CRC and inErrors, although the port the BHM terminated into in the CMM-Micro reported Zero errors. It ended up being a bad cable from Best Tronics. There were tears in the outside jacket AND the shielding inside. Water had made its way into the cable through the outside holes. Error city, but, only on the BH port, not the CMM port.
What about BH 3, 4, 5, 6? What types are those?
My guess would be that 6 is a Slave, 5 is a Master.
Not sure about 3 and 4…could be Masters or Slaves. Depending on what they are, their bands, and how they are getting their sync, they could also be issues.
Keep me posted.
Tomorrow will be run a new cable for BH1, cut over at midnight.
BH 3 Master uptime 25 days 5.7Ghz
inoctects Count 3390445054
inucastpkts Count 1932119221
Innucastpkts Count 172702246
indiscards Count 0
inerrors Count 844
inunknownprotos Count 0
outoctets Count 3998915641
outucastpktsCount 1871164965
outnucastpkts Count 68799086
outdiscards Count 0
outerrors Count 58
RxBabErr 0
TxHbErr 0
EthBusErr 0
CRCError 0
RxOverrun 844
LateCollision 0
RetransLimitExp 0
TxUnderrun 58
CarSenseLost 0
BH 4 uptime 59 days
inoctects Count 4144129014
inucastpkts Count 226318626
Innucastpkts Count 145895629
indiscards Count 0
inerrors Count 1083
inunknownprotos Count 0
outoctets Count 2975659638
outucastpktsCount 227907045
outnucastpkts Count 363841166
outdiscards Count 1528
outerrors Count 96
RxBabErr 0
TxHbErr 0
EthBusErr 0
CRCError 0
RxOverrun 1083
LateCollision 0
RetransLimitExp 0
TxUnderrun 96
CarSenseLost 0
Bh 5 Master 2.4GHz
inoctects Count 252087348
inucastpkts Count 226482075
Innucastpkts Count 360462554
indiscards Count 0
inerrors Count 801
inunknownprotos Count 0
outoctets Count 547643776
outucastpktsCount 225420167
outnucastpkts Count 150308947
outdiscards Count 13907
outerrors Count 62
RxBabErr 0
TxHbErr 0
EthBusErr 0
CRCError 0
RxOverrun 801
LateCollision 0
RetransLimitExp 0
TxUnderrun 62
CarSenseLost 0
BH 6 slave uptime 5 days
inoctects Count 2312427993
inucastpkts Count 350613325
Innucastpkts Count 16532466
indiscards Count 0
inerrors Count 535
inunknownprotos Count 0
outoctets Count 4275199756
outucastpktsCount 365666299
outnucastpkts Count 37193600
outdiscards Count 43
outerrors Count 10
RxBabErr 0
TxHbErr 0
EthBusErr 0
CRCError 0
RxOverrun 535
LateCollision 0
RetransLimitExp 0
TxUnderrun 10
CarSenseLost 0
msmith:
I changed the cable:
Success rate is 100 percent (10000/10000), round-trip min/avg/max = 1/3/56 ms
That looks much better, checking for CRC errors…
gonna continue to go through the rest of the RF changing cables
Nice. Glad to hear. Let me know how the rest goes.