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 wOOdy-Soft on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

Please Help - Global Download Problems

Status
Not open for further replies.

IllegalOperation

Technical User
Jan 27, 2003
206
US
Hello all. I have been troubleshooting this problem for over a week, but I have little to show for. Unfortunately, this problem has been affecting our customers, so I consider this one to be mission critical. I will be forever in debt to the one that shows me the light on this.

First off, a little background on my network. I have a single 7206VXR NPE400/2FE at my main office, that is pretty much doing all the necessary work. It is acting as my NAT box, DHCP server, firewall, time server, point-to-point circuit collector, and WAN router. Why? Because that is all I had in my budget for. The advantage (besides cost effectiveness) is simplicity.

To continue forward, I have three remote complexes - each with about 15 residential internet access customers for now. These remote complexes are tied into the 7206 with point-to-point DS1s. The A side, of course, is the 7206. More specifically, slot 1 (PA-MC-8T1+). The Z side are all 2621s with a WIC-1DSU-T1. The encapsulation used for the links is HDLC. To keep it simple, I use static routing. Distribution beyond these 2621s is irrelevant to this problem.

Now my WAN link, which is the most important to know about. It was cheaper for me to order two full T1s instead of a fractional DS3. So I have two T1s, set up for load balancing with CEF (per destination). These two T1s also terminate into my PA-MC-8T1+ on slot 1 of the chassis, occupying the first two interfaces. Encapsulation method used is frame-relay. Through these two T1s, my provider has routed me a /28 network block to use publically. If I wanted to accomodate all my future customers, I would have needed at least a public /24. Therefore, I am forced to use NAT/PAT. Once again, no dynamic routing protocols are used for these two links.

Problem: Every now and then, about three times an hour on average, my downloading speeds go down the tube. This happens anywhere off of my main office 7206 - whether I am deep in the network or running locally off the device with a laptop. I normally download around 180KB/Sec, but when this problem occurs it drops down to around 14KB/Sec. This problem happens at completely random times, and the longetivity of it is also random. This network became active a couple weeks ago. I tested out the network from A to Z thoroughly before giving the green light, and it showed no signs of performance degradation. Unfortunately, this problem started occurring soon after customers starting to come online. I have discovered no patterns at all. Now some facts...

- To identify the source of this problem, I pretty much ruled out any device other than the 7206. I have done this by shutting down ALL the interfaces except for my two DIAs while the problem occurs. That means all my point-to-point circuits are all DOWN/DOWN, and I do all my testing locally off one of the ethernet interfaces.

- The problem is NOT our WAN circuits being overloaded. When this issue occurs, I have monitored the bandwidth of the circuit closely. The load on the WAN interface rarely jumps beyond 5%. The peak is probably around 15%.

- When the problem occurs, I also check to see if any packets are dropped. I have sent out thousands of packets, without a single one timing out. There is also no latency on ICMP packets while the problem happens at all. From the 7206 router, I can ping yahoo.com at 14ms. From the end device (which is either a laptop or a PC), I can ping yahoo.com at 23ms. Strange, considering I am only downloading files at around 16KB/Sec. All circuits physically check out 100% ok, including the point-to-point circuits. This has been verified by my service provider as well.

- The CPU on my 7206 has never gone above 1%, even while this issue happens. There is no memory loss either.

- Like I said all circuits look perfectly clean, even while this problem occurs. There are no input/output errors on the interfaces, and no carrier transitions either (unless I manually create them obviously). All LMIs have been successfully sent and received. No BECNs or FECNs either. As mentioned earlier, I had all circuits thoroughly tested plenty of times.

- I doubt the possibility of a virus is causing this, because I have the the latest OS versions with the latest anti-virus definitions. Cisco also has not posted any security issues on their website that could be related to what I am experiencing. My 7206 also has the latest flash available. I have taken every device off the WAN routers when doing the testing, which means all I had connected was either a laptop or a PC.

- To verify that this isnt a bandwidth problem, my provider has given me a 7 day performance report for both DIA links, and both look crystal clean. There are no signs of any dropped packets, or any bandwidth overload at any time. I have not opened up a trouble ticket with them, since they do not see any problems on their side.

- I have taken out as much software possibilities as possible with the 7206, and kept my configurations down to the bare minimum that will enable my customers to maintain their internet connection. This includes removing all access lists and routing protocols (I switched over to static routing). I have noticed nothing out of the ordinary when running Cisco's debug features.

- IOS diagnostics tell me that all hardware is functioning properly. There are no signs of hardware failure at all.

- The buffers on all my interfaces look clean, and my queueing shows no signs of overload.....even while the problem is currently happening. I even turned off queueing for the sake of troubleshooting just to make sure.

- Even though my download speeds have been randomly dropping, my upload speeds remain intact. I ran some tests while this problem happens, and I can still upload at pretty much the full 1.544Mbps/Sec......even though I am downloading at only like 15KB/Sec.

That is all off the top of my head right now, but please feel free to ask me some questions. Like I said I have spent a week on this, and I am ready to give up any detail. I know you are going to want to see my config, so I will post that on a different post (this one is getting long). Everything points to my configuration of this router, since it is pretty much the only peice of the puzzle left. Like I said, I will sincerely appreciate everyones effort on this one. Your one smart cookie if you can solve this on the first try (either that or Im pretty dumb for missing something silly). Thanks for the assistance....
 
Here is my config. Dont really worry about the access lists for now, because I have disabled them to troubleshoot this problem (this problem still happens when they are disabled). There are a couple things that concern me. One are my NAT pools. I was forced to do them strange like that, since I have two outbound/WAN interfaces. Notice that there is only one IP address available for each pool (I am doing overload). I have debugged NAT and it appears to be doing ok. Also, for administrative purposes - I assigned a different channel group number for each of my serial interfaces. I always thought that is only locally significant, so I simply assigned one channel on each interface all available time slots. Both ends of my point-to-point circuits also have their clock source set to internal. Ive always been told that works best. Lastly, I have a few IP addresses that are currently "offline" on my NTP configurations. I didnt think doing that would be significant however. Well, here it is...

KEY:
X = Insignificant Information
A = WAN IP Address Information for WAN Circuit #1
B = WAN IP Address Information for WAN Circuit #2
C = First IP Address of Assigned Public Network Block
D = Second IP Address of Assigned Public Network Block
E = Third IP Address of Assigned Public Network Block
Z = Insignificant IP Addresses

Current configuration : 10790 bytes
!
!
version 12.1
service timestamps debug uptime
service timestamps log uptime
service password-encryption
!
hostname XXX
!
card type t1 1
logging rate-limit 10 except errors
no logging console
enable secret XXX
enable password XXX
!
clock timezone EST -5
clock summer-time EDT recurring
ip subnet-zero
no ip source-route
ip cef
!
!
ip name-server X.X.X.X
ip name-server X.X.X.X
no ip dhcp conflict logging
ip dhcp excluded-address 192.168.2.1 192.168.2.9
ip dhcp excluded-address 192.168.3.1 192.168.3.9
ip dhcp excluded-address 192.168.6.1 192.168.6.9
!
ip dhcp pool CLE1-2
network 192.168.2.0 255.255.255.0
dns-server X.X.X.X 206.141.193.55
default-router 192.168.2.1
!
ip dhcp pool CLE1-3
network 192.168.3.0 255.255.255.0
dns-server X.X.X.X 206.141.193.55
default-router 192.168.3.1
!
ip dhcp pool CLE1-6
network 192.168.6.0 255.255.255.0
dns-server X.X.X.X 206.141.193.55
default-router 192.168.6.1
!
no ip bootp server
!
controller T1 1/0
cablelength long 0db
channel-group 0 timeslots 1-24
!
controller T1 1/1
cablelength long 0db
channel-group 1 timeslots 1-24
!
controller T1 1/2
clock source internal
cablelength long 0db
channel-group 2 timeslots 1-24
!
controller T1 1/3
clock source internal
cablelength long 0db
channel-group 3 timeslots 1-24
!
controller T1 1/4
shutdown
cablelength long 0db
channel-group 4 timeslots 1-24
!
controller T1 1/5
shutdown
cablelength long 0db
channel-group 5 timeslots 1-24
!
controller T1 1/6
clock source internal
cablelength long 0db
channel-group 6 timeslots 1-24
!
controller T1 1/7
shutdown
cablelength long 0db
channel-group 7 timeslots 1-24
!
!
interface Port-channel1
no ip address
hold-queue 300 in
!
interface FastEthernet0/0
description LAN Test Interface
shutdown
!
interface FastEthernet0/1
description LAN Interface
ip address 192.168.0.2 255.255.255.0
ip access-group 103 in
ip access-group 104 out
duplex auto
speed auto
no cdp enable
!
interface Serial1/0:0
description WAN Interface #1
no ip address
encapsulation frame-relay IETF
frame-relay lmi-type ansi
!
interface Serial1/0:0.XXX point-to-point
description WAN Link #1
ip address A.A.A.78 255.255.255.252
ip access-group 101 in
ip access-group 102 out
no ip redirects
ip nat outside
no arp frame-relay
no cdp enable
frame-relay interface-dlci XXX IETF
!
interface Serial1/1:1
description WAN Interface #2
no ip address
encapsulation frame-relay IETF
frame-relay lmi-type ansi
!
interface Serial1/1:1.ZZZ point-to-point
description WAN Link #2
ip address B.B.B.74 255.255.255.252
ip access-group 101 in
ip access-group 102 out
no ip redirects
ip nat outside
no arp frame-relay
no cdp enable
frame-relay interface-dlci ZZZ IETF
!
interface Serial1/2:2
description Link to Remote Site #1
ip address 10.0.2.1 255.255.255.252
ip nat inside
!
interface Serial1/3:3
description Link to Remote Site #2
ip address 10.0.3.1 255.255.255.252
ip nat inside
!
interface Serial1/4:4
no ip address
shutdown
!
interface Serial1/5:5
no ip address
shutdown
!
interface Serial1/6:6
description Link to Remote Site #3
ip address 10.0.6.1 255.255.255.252
ip nat inside
!
interface Serial1/7:7
no ip address
shutdown
!
ip nat pool CLE1-2 C.C.C.115 C.C.C.115 prefix-length 28
ip nat pool CLE1-6 D.D.D.114 D.D.D.114 prefix-length 28
ip nat pool CLE1-3 E.E.E.116 E.E.E.116 prefix-length 28
ip nat inside source list 2 pool CLE1-2 overload
ip nat inside source list 3 pool CLE1-3 overload
ip nat inside source list 6 pool CLE1-6 overload
ip classless
ip route 0.0.0.0 0.0.0.0 A.A.A.77
ip route 0.0.0.0 0.0.0.0 B.B.B.73
ip route 192.168.2.0 255.255.255.0 10.0.2.2
ip route 192.168.3.0 255.255.255.0 10.0.3.2
ip route 192.168.6.0 255.255.255.0 10.0.6.2
no ip http server
!
!
logging facility local1
logging 192.168.0.8
! NAT Access Lists
access-list 2 permit 192.168.2.0 0.0.0.255
access-list 3 permit 192.168.3.0 0.0.0.255
access-list 6 permit 192.168.6.0 0.0.0.255
! WAN Interface Access Lists
access-list 101 deny tcp any any eq 0
access-list 101 deny tcp any any eq 69
access-list 101 deny tcp any any eq 135
access-list 101 deny tcp any any eq 137
access-list 101 deny tcp any any eq 138
access-list 101 deny tcp any any eq 139
access-list 101 deny tcp any any eq 161
access-list 101 deny tcp any any eq 162
access-list 101 deny tcp any any eq bgp
access-list 101 deny tcp any any eq 445
access-list 101 deny tcp any any eq 593
access-list 101 deny tcp any any eq 1433
access-list 101 deny tcp any any eq 4444
access-list 101 deny udp any any eq 0
access-list 101 deny udp any any eq tftp
access-list 101 deny udp any any eq netbios-ns
access-list 101 deny udp any any eq netbios-dgm
access-list 101 deny udp any any eq 135
access-list 101 deny udp any any eq netbios-ss
access-list 101 deny udp any any eq snmp
access-list 101 deny udp any any eq snmptrap
access-list 101 deny udp any any eq 179
access-list 101 deny udp any any eq 445
access-list 101 deny udp any any eq 1434
access-list 101 permit ip Z.Z.Z.Z 0.0.0.255 any
access-list 101 permit ip Z.Z.Z.Z 0.0.0.255 any
access-list 101 deny ip any 192.168.0.0 0.0.0.255
access-list 101 deny ip 192.168.0.0 0.0.255.255 any
access-list 101 deny ip any 10.0.0.0 0.0.255.255
access-list 101 deny ip 10.0.0.0 0.0.255.255 any
access-list 101 permit ip any any
access-list 102 deny tcp any any eq telnet
access-list 102 deny tcp any any eq 135
access-list 102 deny tcp any any eq 137
access-list 102 deny tcp any any eq 138
access-list 102 deny tcp any any eq 139
access-list 102 deny tcp any any eq 161
access-list 102 deny tcp any any eq 162
access-list 102 deny tcp any any eq 445
access-list 102 deny tcp any any eq 593
access-list 102 deny tcp any any eq 1433
access-list 102 deny tcp any any eq 4444
access-list 102 deny udp any any eq netbios-ns
access-list 102 deny udp any any eq netbios-dgm
access-list 102 deny udp any any eq tftp
access-list 102 deny udp any any eq 135
access-list 102 deny udp any any eq netbios-ss
access-list 102 deny udp any any eq snmp
access-list 102 deny udp any any eq snmptrap
access-list 102 deny udp any any eq 445
access-list 102 deny udp any any eq 1434
access-list 102 permit ip any Z.Z.Z.Z 0.0.0.255
access-list 102 permit ip any Z.Z.Z.Z 0.0.0.255
access-list 102 deny ip any 192.168.0.0 0.0.255.255
access-list 102 deny ip any 10.0.0.0 0.0.255.255
access-list 102 deny ip 10.0.0.0 0.0.255.255 any
access-list 102 permit ip any any
! LAN Interface Access Lists
access-list 103 deny ip host 192.168.0.1 any
access-list 103 deny ip host 192.168.0.2 any
access-list 103 deny ip host 192.168.0.4 any
access-list 103 deny ip host 192.168.0.44 any
access-list 103 permit ip any host 192.168.2.2
access-list 103 permit ip any host 192.168.3.2
access-list 103 permit ip any host 192.168.6.2
access-list 103 permit ip any host 192.168.0.222
access-list 104 deny ip any host 192.168.0.1
access-list 104 deny ip any host 192.168.0.2
access-list 104 deny ip any host 192.168.0.4
access-list 104 deny ip any host 192.168.0.44
access-list 104 permit ip host 192.168.1.2 any
access-list 104 permit ip host 192.168.2.2 any
access-list 104 permit ip host 192.168.3.2 any
access-list 104 permit ip host 192.168.4.2 any
access-list 104 permit ip host 192.168.5.2 any
access-list 104 permit ip host 192.168.6.2 any
access-list 104 permit ip host 192.168.0.254 any
cdp timer 120
cdp holdtime 200
!
line con 0
stopbits 1
line aux 0
stopbits 1
line vty 0 4
password XXX
login
line vty 5 15
password XXX
login
!
ntp clock-period 17179937
ntp update-calendar
ntp max-associations 9
ntp peer 10.0.2.2
ntp peer 10.0.3.2
ntp peer 10.0.1.2
ntp peer 10.0.6.2
ntp peer 10.0.7.2
ntp server X.X.X.11
ntp peer 10.0.4.2
ntp peer 10.0.5.2
ntp server X.X.X.1
end


Just what I said before, everything runs just perfect except for this random download performance degradation. It seems that this only happens when downloading files and actual websites. ICMP packets pass normally without any latency, and the bandwidth utilization on my interfaces show no signs of overburden....even when this problem happens.

PLEASE feel free to ask me any questions. I will be more than happy to answer. Once again, I appreciate any assistance...
 
I have a link of screen shots to illustrate what I am seeing on my side. The screen shots were created while I was experiencing the problem.....



- The DOS prompt on the far top left corner is pinging yahoo.com from the laptop I was using (which was connected right on to the 7206). I was getting replies around 23ms, which is completely normal.

- The DOS prompt on the top right side was taken from a 2621, which is connected to the Z side of one of my point-to-point DS1s. The A side of this point-to-point DS1 is terminated at my 7206. As you can see, the load over that point-to-point circuit at that time was at a minimal. I took that screen shot to show that the overall bandwidth utilization across my entire network going through my 7206 looks optimal.

- The DOS prompt on the bottom left screen is a screen shot of my WAN router (7206) pinging yahoo.com. This shows that there was no ICMP latency at all.

- The DOS prompt on the bottom right corner shows the statistics of my 7206s serial interface, that the data is traveling through to get to the internet. In other words, that is the last interface on my network before my provider comes into play.

- The file I am downloading was a large file from Microsoft.com. Notice that even though everything on the router (along with ICMP packets) look quite normal, I am only downloading the file at 14.1KB/Sec.
 
Have taken a look at the bandwidth requirements for 7206 . In these routers certain cards take so much bandwidth and you have to figure out what cards you can put where . anything showing in the logs ? Take a look at this link and make sure you have not overloaded one of the buses , looks like you have quite a bit running on this router . Go to
 
Thanks vipergg. Two T1s and three point-to-points are too much for a PA-MC-8T1+ in the first (and only used) slot of a NPE 400? If that is the case, then why do they put eight interfaces on the card? Like I said, there is probably only a total of 15 users right now - and I would imagine there is only a handful on at any given time. I remember using the "builder" utility they had, and it said the chassis should more than handle the stress of the card.

Thanks for the suggestion though, I am going to have to look further into that.

 
I am focusing on the port adapter today, and noticed the following outputs that raised an eyebrow.

This is a show controller of my second WAN interface....


Interface Serial1/1:1 is up
Hardware is PA-MC-8PRI
ROM rev: 0, Firmware rev: 0x0000001E
H/W rev - PCI9054: 12, PM4380: 5
idb=0x6261C5F8, pa=0x620B0C40, pci_base=0x3CC00000, pci_reg_base=0x3CF00000
pa_ssram=0x3CC00000, pa_fpga= 0x3CD00000, freedm=0x3C800000
plx_runtime_reg=0x3CF00040, plx_local_reg=0x3CF00000, plx_dma_reg=0x3CF00080
doorbell=0x3CF00060, framer1=0x3CD10000
int_status_reg=0x3CF00064, init_block=0x3CCFFF00, pci_stats=0x72CE580,
fw_dbg_buf=0x3CC38398, fw_dbg_bss_buf=0x3CC34BCC
fastsend = 0x6049AB44, post_compress = 0x6049AB44, lovenote = 0x0
idb->lc_ip_turbo_fs=0x6054E2FC, ip_routecache=0x11(dfs=disabled/mdfs=disabled)
ds->tx_limited:0 ds->tx_count:0 ds->max_tx_count:20
no alarm present
Timeslot(s) Used:1-24, subrate: 64Kb/s, transmit delay is 0 flags
RxFree queue=0x72F10E0, shadow=0x621112B8
Rx freeq_wt=888, freeq_rd=889, ready_wt=505, ready_rd=504
TxFree queue=0x72F2120, shadow=0x620F128C
Tx freeq_wt=8467, freeq_rd=8467, ready_wt=3695, ready_rd=3694
# of TxFree queue=8191
Freedm fifo (0x620B2760), hp=0x620AC910 indx=182, tp=0x620B0358 indx=2047
reset_count=0 resurrect_count=0
Tx enqueued=8, in_throttle=0, throttled=0, started=16
tx_limited=FALSE
------PA Channel Counters-------
Channel config: map=0xFFFFFF00, subrate=0xFF, crc32=2
Freedm fifo start=0x620AC430, end=0x620AC4F8, num=26
DSX1 state: up, mtu=1524, timeslots=24
tx pkts=0, tx bytes=0, rx pkts=0, rx bytes=0
errs=3633251, orun=0, abort=1854549, exact=1578007


Here is the output of the show controllers T1 command....

T1 1/1 is up.
Applique type is Channelized T1
Cablelength is long gain36 0db
No alarms detected.
Framing is ESF, Line Code is B8ZS, Clock Source is Line.
Data in current interval (574 seconds elapsed):
0 Line Code Violations, 0 Path Code Violations
0 Slip Secs, 0 Fr Loss Secs, 0 Line Err Secs, 0 Degraded Mins
0 Errored Secs, 0 Bursty Err Secs, 0 Severely Err Secs, 0 Unavail Secs
Total Data (last 24 hours)
0 Line Code Violations, 2 Path Code Violations,
0 Slip Secs, 1 Fr Loss Secs, 0 Line Err Secs, 0 Degraded Mins,
1 Errored Secs, 0 Bursty Err Secs, 1 Severely Err Secs, 0 Unavail Secs


If you notice, I am getting a large amount of errors on my channel counters for that specific interface (3633251). It is also telling me that there are one "errored secs" and one "severely err secs". None of my other interfaces have these problems.

Could THIS be the cause of the problems I am experiencing? I wouldnt think so, because I shut down this interface plenty of times for the sake of troubleshooting in the past (and was still experiencing problems). Also - would those controller errors indicate hardware failure, software failure, physical line problems, etc? I am having difficulty finding the meanings of "aborts" and "exacts" on the internet, along with a few of the other information it is giving me. Thanks for the guidance....
 
Did you ever figure out what was wrong?"

Unfortunately no. I classified it as hardware failure, with the PA-MC-8T1+. Instead of terminating the 2 T1 DIAs directly to the port adapter, I have them running into a 1721 instead as the work around. The configurations are basically exactly the same, with the obvious exception of the different interfaces being used. It runs like a champ with the 1721, or shall I say - it runs like it is supposed to.

I am going to RMA the port adapter and will test out my theory when the new replacement arrives. This isnt high on my priority list right now, but I will give you an update when something comes up. Thanks...
 
well, the first thing I think of is that your WAN connections are Frame-Relay. when you look at them, they may no be overloaded, but your circuit is shared with other customers from your ISP. Do you know what your CIR/BIR rates are? even though it is a T1, its a frame T1, and your CIR (committed rate) may only be 256k or something way less than a T1.
So, what happens is another customer of your ISP is using the same frame T1 as you, they do a big download, and your bandwidth goes down to your CIR, which may be really low. Your ISP doesnt see a problem, but they may be basing that on your CIR. I would check to see what that is. Frame relay can be great cost wise, but its not normally a dedicated T1 for you.

Degg
Network Administrator
 
Thanks Degg, another helpful thought.

Actually, the contract is with SBC....and the CIR is listed at 1.544.

Like I said in a previous post, I ruled out SBC because my DIA workaround (1721) is working great. I havent had a single performance issue yet. Same frame relay circuits, same configurations, different hardware (1721 vs PA-MC-8T1+).

Ill let you guys know if something comes up. Thanks...
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor

Back
Top