Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations TouchToneTommy on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!
MAPD and Definity Limits
- Thread starter fyc896
- Start date
wildcard100001
Technical User
225 ASAI messages per second.... I tried to find the document that states this information, but I could not find it.... This is what I've always said as I have read it before.
Thanks,
Wildcard
Thanks,
Wildcard
- Thread starter
- #3
wildcard100001
Technical User
It is both ways... added together...
Somebody else chime in if I am wrong...
Thanks,
Wildcard
Somebody else chime in if I am wrong...
Thanks,
Wildcard
Details are documented in ASAI Technical Reference.
MAPD handles approximately 200 messages per second to/from the application. It is not a hard limit, rather a resource limit so short bursts may exceed, sustained traffic will break down over these rates.
This is a per card/per link rate. If multiple links, combined traffic up to 200 mps per MAPD is safe, with no more than 200 mps per link. This is a GIP (CM transport layer process) limit that is imposed by a message queue buffer. With CM3, it is possible to achieve 720 mps rates over multiple MAPD cards or AEServices CLAN connections.
With AES on 8700, sustained rates of 720 mps over 4+ CLAN are certified by the labs. Further, AES will load balance when multiple CLAN transport links exist. Further, AES allows for a redundant spare CLAN transport link which will pick up messages should one of the CLAN connections fail.
Reality though is that few applications are inefficient enough to demand >200 mps. Cisco ICM tends to push limits dues to their requirement for near real-time agent status, so they use CMS feeds for 18 second historical agent status and supplement with station queries every 2-3 seconds to update agent status.
For most common softphone or recording apps, it would take something like 5-10K active monitors (agents) with 20-30K bhcc rates to reach 200+ mps.
Generally, a message to CM will match with a from as all messages are ACK or NAK by each side. A typical message like Route Request will have an ACK, then a response/ACK for 4 messages per route request. As the Route request might span 2-3 seconds, the mps of a route request is less than 1 message each way. Alert/connect are similar. The most intense events/requests for messages are alerts/connects/end call events on bridged devices, and Conference/Transfer.
One way to cut traffic is to restrain use of bridged appearances on call center/CTI monitored devices.
MAPD handles approximately 200 messages per second to/from the application. It is not a hard limit, rather a resource limit so short bursts may exceed, sustained traffic will break down over these rates.
This is a per card/per link rate. If multiple links, combined traffic up to 200 mps per MAPD is safe, with no more than 200 mps per link. This is a GIP (CM transport layer process) limit that is imposed by a message queue buffer. With CM3, it is possible to achieve 720 mps rates over multiple MAPD cards or AEServices CLAN connections.
With AES on 8700, sustained rates of 720 mps over 4+ CLAN are certified by the labs. Further, AES will load balance when multiple CLAN transport links exist. Further, AES allows for a redundant spare CLAN transport link which will pick up messages should one of the CLAN connections fail.
Reality though is that few applications are inefficient enough to demand >200 mps. Cisco ICM tends to push limits dues to their requirement for near real-time agent status, so they use CMS feeds for 18 second historical agent status and supplement with station queries every 2-3 seconds to update agent status.
For most common softphone or recording apps, it would take something like 5-10K active monitors (agents) with 20-30K bhcc rates to reach 200+ mps.
Generally, a message to CM will match with a from as all messages are ACK or NAK by each side. A typical message like Route Request will have an ACK, then a response/ACK for 4 messages per route request. As the Route request might span 2-3 seconds, the mps of a route request is less than 1 message each way. Alert/connect are similar. The most intense events/requests for messages are alerts/connects/end call events on bridged devices, and Conference/Transfer.
One way to cut traffic is to restrain use of bridged appearances on call center/CTI monitored devices.
- Status
Similar threads