MultiLink PPP Processor Load

Is there a large increase in Processor Load when using MultiLink PPP to bond T1s together while running MPLS? We have been requested to bond circuits for customers, but are reluctant because of the danger of over utilizing the processor on our 7206s which are mostly NPE-400s and some NPE-G1s. Is there a reference somewhere to the increase in Processor Load caused by MPPP?

Hi,
Cisco does not publish quantitative details on the impact of MLPPP on CPU utilisation. So the best option for you is to do some baselining within the lab: start with an unloaded box with two separate T1s, pass traffic over them so that they are both fully utilised and then note CPU utilisation. Then, enable MPPP over these, pass traffic over the multilink interface so that it's fully utilised and then note CPU utilisation. That should give you some idea of the incremental impact of such interfaces. You can always do things such as disable fragmentation when using MLPPP which does reduce CPU utilisation.
Having said that, the NPE-400/NPE-g1s are pretty grunty and should be able to handle a few multilink interfaces. It comes down to how many of these interfaces you are expecting to use...
Hope that helps - pls do rate the post if it does.
Paresh

Similar Messages

Multilink PPP problem

Hi Guys,
i got a situation here and hope someone will be able to help me out. Recently i configured a PPP multilink and am facing some issues after that.
Issue 1:
There seems to be lot of input error on multilink interface (150 errors in 40 mins)
Issue 2:
when any of the multilink member interface goes down (all other links are active), am getting packrt drops and request timed out when i ping my other end multi link interface atleast for some 30 sec , after which my response is ok.
About my network
I have connectivity between site A and Site B which consists of five 64 kbps Leased lines bundled together with Multilink PPP
My configuration is shown below
End A
interface multilink 1
ip address x.x.x.x x.x.x.x
ppp multilink fragment-delay 20
ppp multilink interleave
ppp multilink
multilink-group 1
int Ser 2/0
bandwidth 64
no ip address
no ip redirects
no ip proxy-arp
encapsulation ppp
no ip route-cache
no ip mroute-cache
no fair-queue
ppp multilink
multilink-group 1
int Ser 2/1
bandwidth 64
no ip address
no ip redirects
no ip proxy-arp
encapsulation ppp
no ip route-cache
no ip mroute-cache
no fair-queue
ppp multilink
multilink-group 1
int Ser 2/2
bandwidth 64
no ip address
no ip redirects
no ip proxy-arp
encapsulation ppp
no ip route-cache
no ip mroute-cache
no fair-queue
ppp multilink
multilink-group 1
int Ser 2/3
bandwidth 64
no ip address
no ip redirects
no ip proxy-arp
encapsulation ppp
no ip route-cache
no ip mroute-cache
no fair-queue
ppp multilink
multilink-group 1
int Ser 2/4
bandwidth 64
no ip address
no ip redirects
no ip proxy-arp
encapsulation ppp
no ip route-cache
no ip mroute-cache
no fair-queue
ppp multilink
multilink-group 1
End B
interface multilink 1
ip address ip address x.x.x.x x.x.x.x
ppp multilink fragment-delay 20
ppp multilink interleave
ppp multilink
multilink-group 1
interface S 2/0:1
bandwidth 64
no ip address
no ip redirects
no ip proxy-arp
encapsulation ppp
no ip route-cache
no ip mroute-cache
no fair-queue
ppp multilink
multilink-group 1
interface S 2/0:11
bandwidth 64
no ip address
no ip redirects
no ip proxy-arp
encapsulation ppp
no ip route-cache
no ip mroute-cache
no fair-queue
ppp multilink
multilink-group 1
interface S 2/0:25
bandwidth 64
no ip address
no ip redirects
no ip proxy-arp
encapsulation ppp
no ip route-cache
no ip mroute-cache
no fair-queue
ppp multilink
multilink-group 1
interface S 2/0:4
bandwidth 64
no ip address
no ip redirects
no ip proxy-arp
encapsulation ppp
no ip route-cache
no ip mroute-cache
no fair-queue
ppp multilink
multilink-group 1
interface S 2/0:6
bandwidth 64
no ip address
no ip redirects
no ip proxy-arp
encapsulation ppp
no ip route-cache
no ip mroute-cache
no fair-queue
ppp multilink
multilink-group 1

pls find the output of show command
END A #sh int multilink 5
Multilink5 is up, line protocol is up
Hardware is multilink group interface
Description: ***Multilink to End B ***
Internet address is 192.26.0.162/30
MTU 1500 bytes, BW 320 Kbit, DLY 100000 usec,
reliability 255/255, txload 2/255, rxload 1/255
Encapsulation PPP, loopback not set
Keepalive set (10 sec)
DTR is pulsed for 2 seconds on reset
LCP Open, multilink Open
Open: IPCP, CDPCP
Last input 00:00:00, output never, output hang never
Last clearing of "show interface" counters 19:55:13
Input queue: 0/75/0/0 (size/max/drops/flushes); Total output drops: 0
Queueing strategy: fifo
Output queue: 0/40 (size/max)
5 minute input rate 2000 bits/sec, 4 packets/sec
5 minute output rate 3000 bits/sec, 4 packets/sec
309478 packets input, 58763590 bytes, 0 no buffer
Received 0 broadcasts, 0 runts, 0 giants, 0 throttles
1492 input errors, 0 CRC, 1492 frame, 0 overrun, 0 ignored, 0 abort
290606 packets output, 32728691 bytes, 0 underruns
0 output errors, 0 collisions, 0 interface resets
0 output buffer failures, 0 output buffers swapped out
0 carrier transitions
END A #sh ppp multilink
Multilink5, bundle name is End B
Bundle up for 22:06:31
493 lost fragments, 162050 reordered, 0 unassigned
999 discarded, 0 lost received, 3/255 load
0xAE5E6 received sequence, 0x69F6B sent sequence
Member links: 5 active, 0 inactive (max not set, min not set)
Se2/0, since 20:07:34, last rcvd seq 0AE5E1, 160 weight, 152 frag size
Se2/2, since 20:04:45, last rcvd seq 0AE5E2, 160 weight, 152 frag size
Se2/3, since 20:04:07, last rcvd seq 0AE5E3, 160 weight, 152 frag size
Se2/5, since 20:03:24, last rcvd seq 0AE5E4, 160 weight, 152 frag size
Se2/1, since 19:57:12, last rcvd seq 0AE5E5, 160 weight, 152 frag size
Am getting 1400 input errors in last 19 hours on multilink interface, also less than 100 input errors on member serial interfacs.Also am not loosing any link member.

MPLS TE with Multilink PPP failed

I have a problem setting MPLS TE in multilink PPP.
I setup 2xe1 PE to P connection using multilink PPP. Configuration for MPLS TE had been completely setup.
Tunnel MPLS was terminated at interface multilink at both end and also the rsvp had been attached to these interfaces.
PE1 ==== MPLS CORE --- PE2
The configuration seems like this:
mpls traffic-eng tunnels
interface Tunnel100
ip unnumbered Loopback0
tunnel destination 103.23.4.2 tunnel mode mpls traffic-eng
tunnel mpls traffic-eng autoroute announce
tunnel mpls traffic-eng priority 0 0
tunnel mpls traffic-eng bandwidth 100
tunnel mpls traffic-eng path-option 2 dynamic
tunnel mpls traffic-eng record-route
interface Multilink20
description MultiLink to 7513 Gandul
ip address 103.23.4.13 255.255.255.252
ip ospf authentication-key 7 xxxxx
ip ospf hello-interval 5
ip ospf dead-interval 15
mpls label protocol ldp
mpls traffic-eng tunnels
tag-switching mtu 1512
tag-switching ip
ppp multilink
ppp multilink interleave
no ppp multilink fragmentation
multilink-group 20
ip rsvp bandwidth
end
interface Serial3/0:0
no ip address
ip load-sharing per-packet
encapsulation ppp
fair-queue 64 16 48
ppp multilink
multilink-group 20
interface Serial3/1:0
no ip address
ip load-sharing per-packet
encapsulation ppp
ppp multilink
multilink-group 20
With this config the signalling stay down.
Name: PE-CLG-72_t100 (Tunnel100) Destination: 103.23.4.2
Status:
Admin: up Oper: down Path: not valid Signalling: Down
path option 2, type dynamic
Config Parameters:
Bandwidth: 100 kbps (Global) Priority: 0 0 Affinity: 0x0/0xFFFF
Metric Type: TE (default)
AutoRoute: enabled LockDown: disabled Loadshare: 100 bw-based
auto-bw: disabled
Shortest Unconstrained Path Info:
Path Weight: UNKNOWN
Explicit Route: UNKNOWN
History:
Tunnel:
Time since created: 1 hours, 4 minutes
Path Option 2:
Last Error: PCALC:: Local system ID has not been set
Thnks for any response.

hi ,
i have been troubleshooting the same issue for the last 2 days , i go the same message
*Mar 2 02:06:34.973: TE-PCALC_PATH: get_path: system_id not initialized!
R6#show mpls traffic-eng tunnels
Name: R6_t1                               (Tunnel1) Destination: 1.1.1.1
Status:
    Admin: up         Oper: down   Path: not valid   Signalling: Down
    path option 10, type dynamic
Config Parameters:
    Bandwidth: 0        kbps (Global) Priority: 7 7   Affinity: 0x0/0xFFFF
    Metric Type: TE (default)
    AutoRoute: disabled LockDown: disabled Loadshare: 0        bw-based
    auto-bw: disabled
History:
    Tunnel:
      Time since created: 1 days, 17 minutes
    Path Option 10:
      Last Error: PCALC:: Local system ID has not been set
R6#
i am using ISIS as routing protocol , I dont know if is a bug with GNS3 or something is wrong with my config wich i dont think so !!
can anybody help me please !

Two T1 lines and multilink PPP question

We are getting an additional full T1 line from our ISP for failover and increased bandwith. Each T1 line will go to different NOCs. We decided on multilink ppp (MLPPP). Our ISP told us that traffic will be transmitted on both T1 lines. Packets from one NOC will go to the other NOC, reassembled and send to WAN. In the event one NOC fails, all traffic will go through the other one.
I understand two MLPPP lines must terminate on a single router at a single location. Is what I mentioned above possible?
I also have a 2651XM router with 256MB of memory. I need two seperate WIC CSU/DSU integrated T1 modules. Can someone recommend a model?
Thanks

The most common implementation is indeed MLPPP which terminates multiple PPP links on the same end devices. There is also an implementation of multi-chassis multilink which can terminate ppp links on separate chassis. I guess that is is possible that the chassis could be in separate NOCs. I have no experience with this implementation where the participating routers are in different NOCs, but I assume that it could work.
HTH
Rick

8.01 processor load

The new 8.01 seems to be more stable however the CPU meter runs along great (low load) for about 10 minutes then all of a sudden I get heavy CPU load and the odd cut out. Last night mainstage crashed completely during a gig. I basically using 1 EVB3 patch
Has anyone experience heavy processor load after the software has been active for a 10 to 20 minutes
I'm trying to see if I have a hardware problem or not
JN

I just recognized that all of a sudden Spotlight started indexing iCal files (found out in Konsole:
22.12.07 00:26:58 mdworker[760] iCal.mdimporter: ignoring file </Users/Library/Application Support/iCal/Sources/1F2BE646-18F3-42B2-9D0C-66BA018AC293.calendar/corestorage. ics>)
- and the fans started blowing. Had a look in Activity monitor and the processor load jumped up to 50%.
That would be killing any hard activity with EVB3s.
Maybe there is a way to deactivate Spotlight before a gig.
00:26:58 is the right time to do maintenance for office workers but not for musicians.
That's the time when the crowd should get hot, not the processor.

100% processor load

Hello there,
I checked out a new strange behavior with TestStand 3.1 and LabVIEW 7.1. I use a SubVi in a Step to get the SubPropertiesNames. (See "GetSubProps2.JPG") I want to code it right and used the "Merge Errors.vi", because if the "Close Reference.vi" gets an error the reference will not be closed. But after closing the Step after calling via EditStep-method, the processor load goes up to 100%.
I had hard to search for the cause. I had to remove the "Merge Errors.vi" that it will be work. (see "GetSubProps1.JPG") But why?
I can't understand this strange behavior. So if somebody has a made a similar experience - please answer..
Ciao TPoint
Attachments:
GetSubProps2.JPG ‏59 KB
GetSubProps1.JPG ‏57 KB

Hi TPoint,
It probably is working but it depends how big that error cluster is, it could be quite large, an you know what happens to labview an larger array of clusters can do, especially if they is no delay in a loop. Basically it grabs all the processor.
There was a problem, I think the change from labVIEW 5.x to labVIEW 6.0, (I think) which caused a problem in the LabVIEW Operator Interface. This was resolved by passing the reference through a shift register, as you have done with the error.
Now, I also do this with references.
Regards
Ray Farmer
Regards
Ray Farmer

Parallels vs Fusion and processor load

Parallels causes my MBA to run very hot and causes the fans to run at full speed.
Does anybody know if Fusion has the same effect or does it cause less processor load?

Pagemakers2 wrote:
Can anyone confirm if Fusion applied the same load as Parallels? If it's a Windows thing then I guess it does.
Yes, I can confirm it. In this aspect, there is no difference between Fusion and Parallels. Once Windows settles down and gets updated, it will run much more quietly. Windows will always have a tendency to cause more CPU and fan usage. It's normal.
Also, there was a bug in Fusion (maybe fixed now?) that caused the airport process on the Mac to spike to 100%, whether Windows was running or not. That is a different issue. Parallels vs. Fusion is somewhat of a religious war. People like one or the other. When it comes to Windows CPU and fan usage, they will be the same.
I know it makes my MBA warm and causes the fans to run at full tilt. Will it also do the same thing on an iMac?
Yes. You don't really have to worry about it though. Any process that causes high CPU usage will do that. The fact that the fans come on is a good thing and means that the Mac has control of things and won't let it go too far.

Sync-server process running with about 30% processor load

Hi there !
After installing 10.8.2 I have this issue.
Can anybody explain to me, what is going on with this process.
Does it stop or is there anything I must do ?
Thanks in advance !

I had the same symptoms and found a solution. I went to System Preferences>iCloud. I clicked "sign out". I agreed to all of the windows that poped up. Then I restarted my laptop. Then I went back to System Preferences to sign back into all of my iCloud services. After this the issue went away. It seems like Contacts was constantly trying to sync something, but reseting my iCloud acccount on my laptop seemed to stop SyncServer from running at 20% of my processor load 24/7.

Is it possible my FieldPoint's clock is getting extended due to processor load?

I am trying to measure a signal that is around 10 Hz using the cFP-AIO-600. It is a basic pulse counter .vi that I am uploading into the cFP-2020 that counts all pulses that occur in one second . The odd thing that I am seeing, is that it is counting more pulses than I expect. Instead of 10 pulses per second, it is counting 32. The timestamps of each datapoint "verify" that this occurs in one second.
When I add more functionality (by adding a Collector module and a waveform graph), it somehow counts even more pulses for the same frequency signal. Up to 347 pulses!
How is that possible? It can only be one of two things:
1) The signal actually got faster in between the two measurements (unlikely).
2) Additional .vi code on the cFP causes the clock to get confused, and the 'second' that I am measuring is actually longer than a second by a significant margin.
I have verified this several times, running the second scenario right after the first scenario. The run time of these .vi's take longer than a second (and the .vi with the Collector does seem to take about 10x as long) but this evidence is not as supportive to my theory due to the time introduced by my computer and the cFP communicating with each other and updating graphics.
I have attached these two examples. You will not be able to run them yourself (because of the data access form the cFP-AIO unit) but I have made the results that I got the default values. You should be able to see both my code and the results I got illustrated above when you open the .vi's. pulse_problem_1 shows the first scenario, and pulse_problem_2 shows the second, which is the same thing as pulse_problem_1 with only the Collector and waveform display added.
What can I do to resolve this? It's only a 10 Hz wave, and I know cFP can count even faster signals, so something in my .vi must be wrong. Any help or advice you can give me would be appreciated.
Thanks.
Attachments:
pulse_problem_1.vi ‏123 KB
pulse_problem_2.vi ‏189 KB

Okay, I have proof that my FieldPoint is lying to me when I ask it to do 1000 iterations of 1 wait-until-next-ms loops.
I started with the very useful example in the RT package called "NI Timestamp Code Timer" which, I can only assume, calculates the known timestamp discrepancy caused by processor load. My attached picture (talyor_loop_vs_time_discrepancy2.JPG) shows that right before the cFP-AIO-600 module records a positive voltage, the normally 1 ms loop can take as long as 60 ms. That's fine, but the FieldPoint then reports that only 1 ms has occurred! As you can imagine, this has made pulse counting very difficult, for instead of measuring 10 pulses per second, it 'thinks' it is counting ~28-30 per second as each pulse causes a measurement delay that the FieldPoint treats as a regular 1 ms time interval.
I have also attached the Loop Test code, and a picture of the signal through an ocilliscope, proving it is a regular square wave.
Why does this occur with positive voltages only? Why doesn't it take the same amount of time to read voltages equal to zero? If the delay was constant on every read, I could compensate for it, but I can't compensate for this delay that only occurs on positive pulses!
Please help!
Attachments:
Loop Test.vi ‏126 KB
oscilliscope_measured_taylor.JPG ‏1872 KB
talyor_loop_vs_time_discrepancy2.JPG ‏118 KB

High processor load slows down wireless connection

My wireless connection slows to a crawl when the CPU is being utilized more than 80% (average for both CPU cores).
For example, if I have Safari and Quicktime open, web pages load very fast and the CPU load is less than 5%. When encoding a video in Quicktime, the processor load is around 85%, but web pages load very slowly in Safari (or any other browser). Ping times also increase considerably (from 25ms under low CPU load to 200-2000ms under heavy CPU load). As soon as the video is done encoding and the CPU load drops back down, web pages load fast again.
This problems only seems to apply to AirPort connections. High CPU load does not appear to affect speeds when connected via Ethernet.
Has anyone else experienced this?
MacBook Pro 15" 2.33GHz/2GB/160GB/256MB Mac OS X (10.4.9)

Has no one else experienced this? I can reproduce this problem consistently. Anything that utilizes the processor at least 40% slows down my wireless internet connection.

RAS access via multilink PPP

I need to establish RAS access to my network via an ISDN PRI line. Most users will be single POTS line users but I'll also need to allow bundled POTS line users via Microsoft multilink PPP. I'm planning to order a Cisco 2811 to handle this using a NM-1CE1T1-PRI module for the PRI and a PVDM2-24DM for the calls. Will this setup handle my requirments?

I have 20 leased line with 256 kpbs sharing on 2 E1 links on the module NM-1CE1T1-PRI (1-Port Channelized E1/T1/ISDN-PRI Network Module).
Now we are migrating those 20 line to 2 mbps leased line, i wanted to know that using PVDM can i used the existing 20 lines as backup isdn line ?
We use RAS for dialup and want to remove RAS device so my question is can i use the cards mentioned above so that 20 locations can dial in with on the present E1 card without using patton RAS?

RV320 and Multilink PPP

Has anyone attempted to achieve something like this:
On the remote site, a Cisco RV320 Router, connecting to two DrayTek Vigor120 xDSL modems through the two Gigabit WAN ports, which connect to two xDSL links.
Ignoring the Core end of the xDSL (which I can make support Multilink PPP), has anyone ever imagined/tried/succeeded to run multilink PPP originating in the RV320 router or similar ( the LAC), transiting the two xDSL links and terminating on the LNS at the Core end of the xDSL link?
Sorry its a bit off the wall, I think I may be asking a lot to make this work, but you have to try!
Thanks,
Jim

Bump!...anyone out there with ideas?
Jim

Transparent bridging & Multilink PPP

Is it possible to run Multilink PPP on 2 T1s while also doing transparent bridging?
If so, Do I need to put bridge-group 1 on my multilink interface?
thx

It is not clear what the original problem was or what was the solution. And I am not clear what problem you are trying to solve that leads to the question of bridging over the multi link.
If we knew more about your situation we might be able to give better advice. But I would suggest that the answer will turn out to be something close to this:
- if you want to route IP over the multilink and to bridge some other traffic that might be possible.
- if you want to bridge IP over the multilink then you can not have an IP address configured on the multilink. To bridge IP over the multilink you might need to do something like configure Integrated Routing and Bridging. In IRB you configure a virtual interface BVI and put the IP address on the BVI. then you put a bridge group on the multilink which points to the BVI. It is ugly and I would not recommend it, but I think that it might work.
HTH
Rick

Multilink-PPP

Hi,
i want to make my Archlinux work as Multilink-PPP server.
i played with MPD http://mpd.sourceforge.net/, a FreeBSD Multilink-PPP daemon, and it was easy to let it bundle to L2TP-Connections to a single link.
But doing the same with Linux seems to be quite harder, can you help me?
The-Master

the only way that i know to set up a multilink ppp in linux is to use pppd and set the mp option in the peer file, you can probably find a lot more about it by checking the pppd man pages and you will also need to find a l2tp module for the kernel. search google for it.

Multilink PPP (MLPPP)?

I have a need for multilink PPP. I can't seem to find a way to configure this in my AEBS - is it possible?

The most common implementation is indeed MLPPP which terminates multiple PPP links on the same end devices. There is also an implementation of multi-chassis multilink which can terminate ppp links on separate chassis. I guess that is is possible that the chassis could be in separate NOCs. I have no experience with this implementation where the participating routers are in different NOCs, but I assume that it could work.
HTH
Rick

MultiLink PPP Processor Load

Similar Messages

Maybe you are looking for