mtracer should have a timestamp

trjm · Post by **trjm** » 22 Jan 2010 08:49

When I log ISDN messages over a trunk group, it should be possible to know when they where sent.
If a customer say that problem occured 10:37 on friday, I don't know how to find 10:37 in the output.
If every message had a time stamp that said 201001221037 or something like that, it would be easier.

Rgds
Terje Melberg

Elandryl · Post by **Elandryl** » 22 Jan 2010 09:14

There is a "timestamp"

you got the start time when you launch mtracer :
mtracer started ...
(761779:000001) MTRACER host (192.168.126.6, partitio-cs1), version: R9.0-h1.301-28-b-fr-c0
(761779:000001) MTRACER num: 457, time: 2010/01/11 18:59:26, loss: 0%

After that in every message you got a timer
______________________________________________________________________________
| (761818:000002) 1321: Send_IO1 (link-nbr=1, sapi=0, tei=0) :
| long: 54 desti: 0 source: 15 cryst: 1 cpl: 4 us: 8 term: 1 type a5
| tei: 0 <<<< message sent : SETUP [05] Call ref : 54 a1
|______________________________________________________________________________

______________________________________________________________________________
| (761821:000004) Concatenated-Physical-Event :
| long: 23 desti: 0 source: 0 cryst: 1 cpl: 4 us: 0 term: 1 type a5
| tei: 0 >>>> message received : CALL PROC (02) Call ref : d4 a1
|______________________________________________________________________________
|
| IE:[18] CHANNEL (l=3) a9 83 82 -> T2 : B channel 2 exclusive
|______________________________________________________________________________

761821:000004 - 761818:000002 = 2

This unit is in 100ms just like timer in OXE.
So there is 0.2s between the two paquet.

You have the same timer in the mtracer start informations, so from that you can find the hour or the elapsed time between two paquets as you want.

I know this is not so clear and easy but is here !

Hope it can help you

Enjoy

Post by **frank** » 27 Jan 2010 00:29

Also, when you start mtracer:

mtracer started ...
(362887:000001) MTRACER host (192.168.1.100, lab9), version: R9.0-h1.301-34-us-c0s1
(362887:000001) MTRACER num: 007, time: 2010/01/26 21:27:38, loss: 0%
(362888:000002) get_neqt_country() : country of neqt 30819 = USA(93)
(362888:000003) cstaRunSupFaisceau
(362888:000004) cstaRunSupFaisceau deviceMcdu=7050
(362888:000005) cstaTrunkGroupSup started
(362888:000006) cstaRunSupFaisceau OK
(362888:000007) cstaRunSupFaisceau with timer
(362888:000008) SBC4>> SORTIE de ABCA_Main first
(362895:000009) morcel_basculetat-->nb_paquet=6,nlogass=4
(362895:000010) kbxdate -> date.ordjanne 25
(362895:000011) kbxdate -> date.heure 21
(362895:000012) kbxdate -> date.minute 27
(362895:000013) kbxdate -> date.second 39
(362895:000014) kbxdate -> date.jour 26
(362895:000015) kbxdate -> date.mois 1
(362895:000016) kbxdate -> date.annee 110
(362895:000017) basculetat heure_cour 4167
(362895:000018) basculetat jour_sem 2 (dim =0)

You have the date/time of start.
The second part of all messages (xxx:yyy) yyy gives you how many millisecond elapsed since the start.

Try tuner -s too.. I think this option throw a date/time once in a while.

trjm · Post by **trjm** » 05 Feb 2010 05:16

Not quite I think. TC0672 have this to say

Standard format of monitel messages
(HHHHHH:NNNNNN) < message content >
HHHHHH is a clock relating to the initialization of the telephone application belonging to monitel,
the variations of which are given in steps of 100 ms. Observing this counter gives a rough idea of
the time that passes between several trace messages (the date of a message is fixed when the
message is sent, and not when it is printed).
NNNNNN is a series number, which is increased by one unit each time a new trace request is
made.

So one could in theory find out when the monitel application was started, and count from there. Still, would be better if OXE came with more field engineer friendly tools

Post by **vad** » 05 Feb 2010 06:45

For ISDN trace you can use f.e. t3 with option -data, like this
<- (16:43:16) cr:0 cpl:8 access:0 CR:5695 ### ETABLISSEMENT
00.00.34.00.00.0f.00.08.08.00.a5.00.00.08.02.56. ..4............V
95.05.04.03.80.90.a3.6c.0c.21.81.33.34.32.32.31. .......l.!.34221
30.34.30.31.30.70.08.a1.32.31.30.34.30.30.30.7d. 04010p..2104000}
02.91.81.a1. ....

-> (16:43:17) cr:0 cpl:8 access:0 CR:d695 ### APPEL EN COURS
08.05.17.00.00.00.00.08.00.00.a5.00.00.08.02.d6. ................
95.02.18.03.a9.83.82. .......

-> (16:43:18) cr:0 cpl:8 access:0 CR:d695 ### ALERTE
08.06.1a.00.00.00.00.08.00.00.a5.00.00.08.02.d6. ................
95.01.1e.02.82.88.1e.02.80.88. ..........

<- (16:43:19) cr:0 cpl:8 access:0 CR:5695 ### DECONNEXION
00.00.1a.00.00.0f.00.08.08.00.a5.00.00.08.02.56. ...............V
95.45.08.02.81.90.1e.02.81.88. .E........

-> (16:43:19) cr:0 cpl:8 access:0 CR:d695 ### LIBERATION
08.05.17.00.00.00.00.08.00.00.a5.00.00.08.02.d6. ................
95.4d.08.03.82.e3.1e. .M.....

<- (16:43:19) cr:0 cpl:8 access:0 CR:5695 ### FIN DE LIBERATION
00.00.16.00.00.0f.00.08.08.00.a5.00.00.08.02.56. ...............V
95.5a.08.02.81.90. .Z....

krzysioD · Post by **krzysioD** » 09 May 2010 17:54

Guys, just some bash-izm:

while :; do date; sleep 10 ; done &
mtracer -a

and every 10 s you will get output of command 'date', you could step down towards 1s, but it seems like fine granularity.