EAGLE fully configured for SS7-over-IP consists of at least one IPSG application. The application can be installed on an E5-ENET-B card or a SLIC card.

Table 6-1 shows the cards and their Advertised Capacity in TPS.

The capacities listed in this table are achieved when the traffic carried by the application involves no feature or network attribute that requires excessive CPU, memory, or transport capacity. Rates in excess of the values shown results in signaling link or IP connection congestion.

When monitoring the IPx links using IMF, Oracle requires that the HIPR2 cards and at least one STC card are configured on the same shelf as the IPSG cards. The M2PA and M3UA links that are RFC 4165 compliant can be monitored. A minimum of two STC cards are required per system to turn on the monitoring feature in EAGLE.

The E5IS Data Feed or monitoring subsystem requires a significant amount of CPU and memory resources from the IPSG cards when monitoring the M2PA, M3UA links. When enabled, this capability causes the IPSG applications to drop well below the maximum capacity of the platform. For a detailed analysis of the IP7 throughput for provisioning purposes, refer to Engineering Rules for Determining IP7 Application Throughput.

The installation of the SS7-over-IP system includes both hardware installation and software provisioning, and is detailed in the EAGLE customer documentation.

1. Declare the E5-ENET-B card or the SLIC application to be ipsg (ent-card).
2. Define the IP settings for the Ethernet port (chg-ip-lnk):
   1. Declare what card and port you are defining with this command.
   2. Associate an IP address to that card and port.
   3. Set the Ethernet settings for the card and port.
3. Associate an IP address to a host name used in configuring the association (ent-iphost).
   This step sets up a static IP address Host Table, which associates Domain Names to IP addresses so that the computer can look up Domain Names and place the corresponding IP address in the packet header. The alternative is to use a DNS server.
4. Enter an Application Server Process and bind an SCTP association with it (ent-assoc).
   This command configures the SCTP association in the Internet Protocol Application Socket (IPAPSOCK) table. This command permits the association to transport protocol data units and adaptive layer peer messages. Each association is connected to a process at the far end. The IPAPSOCK table is used to associate the Local Host or Local Port to a Remote Host or Remote Port.
5. Define the Site ID (chg-sid).
6. Enter the adjacent point code (ent-dstn).
   
   
7. Define the capacity and use the alarm (ent-ls).
   ent-ls:lsn=ls1201:apc=10-10-10:lst=a:adapter=m3ua:ipsg=yes:rcontext=1:slktps=100
   
8. Tell EAGLE that this is a SIGTRAN M3UA link (ent-slk).
9. Enter the route (ent-rte).
   
10. Allow and open the SCTP association (chg-assoc).
11. Activate the signaling link (act-slk).

1. Enter an Application Server Process and bind an SCTP association with it (ent-assoc).
2. Enter the adjacent point code (ent-dstn).
3. Define capacity and use the alarm (ent-ls).
4. Tell EAGLE that this is a SIGTRAN M2PA link (ent-slk).
5. Enter the route (ent-rte).
   
6. Allow and open the SCTP association (chg-assoc).
7. Activate the signaling link (act-slk).

Set the RTIMES parameter such that an association will be marked unavailable after a reasonable amount of time, based on the values of the RMODE, RMIN and RMAX parameters.

For M2PA, this should be just after M2PA T7 expires (default 1.2 sec).

For example, consider a unihomed M2PA link with RMIN set to 100 msec and RMODE is LINEAR:

Time to mark as failed = RMIN * RTIMES 1200 msec = 100 msec * 12

As long as RTIMES = 12, the association will fail at about the same time MTP3 starts changeover procedures (12 is the maximum for RTIMES).

In this case, decrease M2PA T7 slightly using the chg-m2pa-tset command to guarantee that it will expire before the association is taken down.

For M3UA connections, make this a reasonable amount of time for the network, remembering that multihomed associations could be taken down after only RTIMES/2 retransmits.

Use the ping-result average RTT measurement for calculation of RMIN.

RMIN should be set to whichever is greater of 1.2 * (Avg. RTT) or (Avg. RTT) + 10 ms.

If errors are greater than 1 per 250,000, then investigate to determine if this can be improved in the network.

RMAX can be set to the worst recorded RTT and further tuned after the association has been established and assocrtt measured.

:rtxthr –The retransmission threshold for the association. The RTXTHR parameter value indicates the number of packet re-transmissions that can occur on the association (per monitoring time period of two seconds). Alarm "IP Connection Excess Retransmits" (UAM 536) will be raised if the number of packets re-transmitted is greater than the configured the RTXTHR parameter value, during five such consecutive monitoring periods. Once alarm is raised, it may require up to 12 consecutive monitoring periods with the number of re-transmissions < RTXTHR to clear the alarm. The design allows the alarm to come on at low error rates, and not come for occasional errors.

The value of this parameter is 0 to 65,535. The value of this parameter is shown in the RTXTHR field of the rtrv-assoc:aname=<association name> output. The rtxthr parameter value can be changed if the open parameter value is either "yes" or "no". It is possible to configure the RTXTHR so that UAM 536 alarms if the error rate on association is above the recommended maximum packet loss of 0.025%. If the error rate is more than 0.025%, investigate to determine if this can be improved in the network.

Measure jitter using ping samples taken from the network. Ideally, a relatively small subset of the samples deviate from the overall Average RTT for the network. The SCTP RMIN parameter value should be adjusted during deployment such that RMIN is approximately equal to 1.2 * Average RTT time in the network. RTT in the network should not exceed 120 ms for the E5-ENET-B cards, or 50 ms for the E5-ENET-B cards running the IPSG application when the E5-ENET-B IPSG High Throughput feature is turned on.

After an association is established, the EAGLE pass command should be used to get the true RTT as experienced by the association.

1. Reset the counters: pass:loc=XXXX:cmd=“assocrtt –r <assoc name>".
2. Wait a reasonable interval (preferably 24 hours) before collecting the measurements: pass:loc=XXXX:cmd=“assocrtt <assoc name>".
3. Perform the sctp –g pegs or sctp –a assocname command to determine if any retransmissions have occurred.
4. Use the values reported to further tune RMIN and RMAX. Use the Weighted Average RTT in this case for defining RMIN.

   ;
   pass:loc=1105:cmd="assocrtt c7000"
   
   Command Accepted - Processing
   
   rlghncxa03w 00-01-27 08:10:00 EST EAGLE5 31.6.0
   pass:loc=1105:cmd="assocrtt c7000"
   Command entered at terminal #1
   
   rlghncxa03w 00-01-27 08:10:00 EST EAGLE5 31.6.0
   PASS: Command sent to card
   
   rlghncxa03w 00-01-27 08:10:00 EST EAGLE5 31.6.0
   
   ASSOCRTT: Association round trip time report (in milliseconds)
   
   Retransmission Configuration
   Retransmission Mode : LIN
   Minimum RTO : 120
   Maximum RTO : 800
   
   Traffic Round-Trip Times
   
   Minimum round-trip time : 5
   Maximum round-trip time : 120
   Weighted Average round-trip time : 10
   Last recorded round-trip time : 10
   
   Measured Congested Traffic Round-Trip Times
   
   Minimum round-trip time : 0
   Maximum round-trip time : 0
   Weighted Average round-trip time : 0
   Last recorded round-trip time : 0
   
   rlghncxa03w 00-01-27 08:10:00 EST EAGLE5 31.6.0
   ASSOCRTT command complete

1. Is the SLIC IPSG card IS-NR (In-service Normal)?

   rept-stat-card:mode=full:loc=<IP CARD location>

2. Is the Ethernet port up or down?

   rept-stat-card:mode=full:loc=<IP CARD location>

3. Are there errors on the Ethernet Interfaces? Are there collisions? CRC errors? Alignment errors? Retransmits?

   pass:loc=<IP card location>:cmd=netstat -d 0 <For Ethernet Interface A>

   pass:loc=<IP card location>:cmd=netstat -d 1 <For Ethernet Interface B>

4. Are there checksum errors?

   pass:loc=<IP card location>:cmd="netstat -p sctp

   Change the SCTP checksum if there are errors, rtrv-sg-opts will show you what checksum is set at; this must match on both ends.

5. Is the far end reachable? Does ping or traceroute work? Is the RTT acceptable? Is there Packet loss?

   pass:loc=<IP card location>:cmd=ping <far-end IP address>

   pass:loc=<IP card location>:cmd="traceroute <far-end IP Address>"

6. What is the delay or jitter of the network?

   pass:loc=<IP card location>:cmd="assocrtt <association>"

7. What is the far end advertising?

   pass:loc=<IP card location>:cmd="sctp -a association>"