Hi All, We have 1 hyperarc that seems to have a problem where it can not release IPs previously assigned properly. It sometimes does and sometimes doesn't ... eventually the arc runs out of IP addresses to assign and spews the following error: Aug 15 09:02:12 9L102PDI-H1 At 13:01:46, Facility "IP", Level "CRITICAL":: ip_fwd_get_opt: no IP address available for dynamic address assignment As you can see, the IP addresses (162 of them) are completely used up and there are only 7 users online. We don't think this is a code specific issue as we run this code on all ARCs (about 250 of them) and this is the only one getting these errors ... it may be hardware related, but in my experience these types of issues are generally code ... so, I'm at a cross-roads so to speak. Has anyone else run into this and found a workaround? BTW: we can find only 2 ways to correct the issue, rebooting or removing the IP Pools and adding them back ... we generally reboot as that is much quicker and less error prone. System Version: V5.0.83/Non-Encr 9L102PDI-H1> list ip pools IP ADDRESS POOLS Name Address Size InUse State Route Unused Priority Status 9L102-H1-POOL1 216.123.143.17/29 6 6 PUBLIC AGGREGATE 0 1 ACTIVE 9L102-H1-POOL2 216.123.140.65/27 30 30 PUBLIC AGGREGATE 0 1 ACTIVE 9L102-H1-POOL3 216.123.132.1/25 126 126 PUBLIC AGGREGATE 0 1 ACTIVE 9L102PDI-H1> list conNECTIONS CONNECTIONS Start Start IfName User Name Type DLL Date Time slot:1/mod:4 xxxxxx DIALIN PPP 14-AUG-2002 21:14:41 slot:1/mod:5 xxxxxx DIALIN PPP 15-AUG-2002 05:13:22 slot:1/mod:11 xxxxxx DIALIN PPP 15-AUG-2002 13:55:43 slot:1/mod:13 xxxxxx DIALIN PPP 15-AUG-2002 03:28:04 slot:1/mod:16 xxxxxx DIALIN PPP 15-AUG-2002 14:04:17 slot:1/mod:19 xxxxxx DIALIN PPP 15-AUG-2002 10:19:32 slot:1/mod:21 xxxxxx DIALIN PPP 15-AUG-2002 08:59:12 9L102PDI-H1> Thanks, James MacDonald, CCNP Senior Platform Manager, IP Network Operations AT&T Canada Corp. 370 King St W., 6th Floor Toronto, ON M5V1J9 Ph: 416-204-5793 Fx: 416-204-5744 Emergency: 416-204-5700 (NMC Hotline) Email: james.macdonald@attcanada.com
Do you have 'hint assigned' enabled? -V ----- Original Message ----- From: "MacDonald, James" <James.MacDonald@attcanada.com> To: <usr-tc@mailman.xmission.com> Sent: Thursday, August 15, 2002 9:39 AM Subject: [USR-TC] ARC not releasing IP addresses
Hi All,
We have 1 hyperarc that seems to have a problem where it can not release IPs previously assigned properly. It sometimes does and sometimes doesn't ... eventually the arc runs out of IP addresses to assign and spews the following error:
Aug 15 09:02:12 9L102PDI-H1 At 13:01:46, Facility "IP", Level "CRITICAL":: ip_fwd_get_opt: no IP address available for dynamic address assignment
As you can see, the IP addresses (162 of them) are completely used up and there are only 7 users online. We don't think this is a code specific issue as we run this code on all ARCs (about 250 of them) and this is the only one getting these errors ... it may be hardware related, but in my experience these types of issues are generally code ... so, I'm at a cross-roads so to speak. Has anyone else run into this and found a workaround?
BTW: we can find only 2 ways to correct the issue, rebooting or removing the IP Pools and adding them back ... we generally reboot as that is much quicker and less error prone.
System Version: V5.0.83/Non-Encr
9L102PDI-H1> list ip pools
IP ADDRESS POOLS Name Address Size InUse State Route Unused Priority Status 9L102-H1-POOL1 216.123.143.17/29 6 6 PUBLIC AGGREGATE 0 1 ACTIVE 9L102-H1-POOL2 216.123.140.65/27 30 30 PUBLIC AGGREGATE 0 1 ACTIVE 9L102-H1-POOL3 216.123.132.1/25 126 126 PUBLIC AGGREGATE 0 1 ACTIVE 9L102PDI-H1> list conNECTIONS
CONNECTIONS
Start Start IfName User Name Type DLL Date Time slot:1/mod:4 xxxxxx DIALIN PPP 14-AUG-2002 21:14:41
slot:1/mod:5 xxxxxx DIALIN PPP 15-AUG-2002 05:13:22 slot:1/mod:11 xxxxxx DIALIN PPP 15-AUG-2002 13:55:43 slot:1/mod:13 xxxxxx DIALIN PPP 15-AUG-2002 03:28:04 slot:1/mod:16 xxxxxx DIALIN PPP 15-AUG-2002 14:04:17 slot:1/mod:19 xxxxxx DIALIN PPP 15-AUG-2002 10:19:32 slot:1/mod:21 xxxxxx DIALIN PPP 15-AUG-2002 08:59:12 9L102PDI-H1>
Thanks,
James MacDonald, CCNP Senior Platform Manager, IP Network Operations AT&T Canada Corp. 370 King St W., 6th Floor Toronto, ON M5V1J9 Ph: 416-204-5793 Fx: 416-204-5744 Emergency: 416-204-5700 (NMC Hotline) Email: james.macdonald@attcanada.com
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
I'm running 3 to 1 now, where have other people started seeing problems at?
I have many chassis setup dual arc (7 dsp's per arc on each chassis). I have not noticed any issues with this. Guess it depends on how much stuff the arc is doing and how much memory it has. You can definitely go more then 3 though. Todd -----Original Message----- From: usr-tc-admin@mailman.xmission.com [mailto:usr-tc-admin@mailman.xmission.com]On Behalf Of Aaron Nabil Sent: Friday, August 16, 2002 11:19 AM To: 'usr-tc@mailman.xmission.com' Subject: [USR-TC] What is a conservative dsp to arc ratio? I'm running 3 to 1 now, where have other people started seeing problems at? _______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
At how many ARCs and DSPs would we need to switch to 130A power supplies? Would having two 70A supplies serve the same purpose as having one 130A, other than the obvious fact that if one of them were to go down, the chassis would be very underpowered. :) Thanks! Chris Kalin ----- Original Message ----- From: "Todd Bertolozzi" <berto@voyager.net> To: <usr-tc@mailman.xmission.com> Sent: Friday, August 16, 2002 11:54 AM Subject: RE: [USR-TC] What is a conservative dsp to arc ratio?
I have many chassis setup dual arc (7 dsp's per arc on each chassis).
I have not noticed any issues with this. Guess it depends on how much stuff the arc is doing and how much memory it has. You can definitely go more then 3 though.
Todd
-----Original Message----- From: usr-tc-admin@mailman.xmission.com [mailto:usr-tc-admin@mailman.xmission.com]On Behalf Of Aaron Nabil Sent: Friday, August 16, 2002 11:19 AM To: 'usr-tc@mailman.xmission.com' Subject: [USR-TC] What is a conservative dsp to arc ratio?
I'm running 3 to 1 now, where have other people started seeing problems at?
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
The following are 3com's recommended specs for maintaining redundancy while running dual power supplies units or PSU's. If you are running 2 PSU's with the following recommendations you will not have any issues if one of them fails. 45 Amp = 6 DSP's, 1 HiperArc, and 1 NMC 70 Amp = 10 DSP's, 1 HiperArc, and 1 NMC 130 Amp = You can fully load the chassis These are what 3com recommends....you can go a bit more but then you're not doing what the equipment vendor recommends. Todd -----Original Message----- From: usr-tc-admin@mailman.xmission.com [mailto:usr-tc-admin@mailman.xmission.com]On Behalf Of Chris A. Kalin Sent: Friday, August 16, 2002 1:22 PM To: usr-tc@mailman.xmission.com Subject: Re: [USR-TC] What is a conservative dsp to arc ratio? At how many ARCs and DSPs would we need to switch to 130A power supplies? Would having two 70A supplies serve the same purpose as having one 130A, other than the obvious fact that if one of them were to go down, the chassis would be very underpowered. :) Thanks! Chris Kalin ----- Original Message ----- From: "Todd Bertolozzi" <berto@voyager.net> To: <usr-tc@mailman.xmission.com> Sent: Friday, August 16, 2002 11:54 AM Subject: RE: [USR-TC] What is a conservative dsp to arc ratio?
I have many chassis setup dual arc (7 dsp's per arc on each chassis).
I have not noticed any issues with this. Guess it depends on how much stuff the arc is doing and how much memory it has. You can definitely go more then 3 though.
Todd
-----Original Message----- From: usr-tc-admin@mailman.xmission.com [mailto:usr-tc-admin@mailman.xmission.com]On Behalf Of Aaron Nabil Sent: Friday, August 16, 2002 11:19 AM To: 'usr-tc@mailman.xmission.com' Subject: [USR-TC] What is a conservative dsp to arc ratio?
I'm running 3 to 1 now, where have other people started seeing problems at?
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
If you are running a full chassis 7 to 1 is a good number, but I have customers who do not want to use more then 1 arc, and they run 10 to 1 with 128Meg on the ARc, seems to work fine it all depends on the RAM, and what the arc is going to be doing. Tom -----Original Message----- From: usr-tc-admin@mailman.xmission.com [mailto:usr-tc-admin@mailman.xmission.com]On Behalf Of Todd Bertolozzi Sent: Friday, August 16, 2002 11:55 AM To: usr-tc@mailman.xmission.com Subject: RE: [USR-TC] What is a conservative dsp to arc ratio? I have many chassis setup dual arc (7 dsp's per arc on each chassis). I have not noticed any issues with this. Guess it depends on how much stuff the arc is doing and how much memory it has. You can definitely go more then 3 though. Todd -----Original Message----- From: usr-tc-admin@mailman.xmission.com [mailto:usr-tc-admin@mailman.xmission.com]On Behalf Of Aaron Nabil Sent: Friday, August 16, 2002 11:19 AM To: 'usr-tc@mailman.xmission.com' Subject: [USR-TC] What is a conservative dsp to arc ratio? I'm running 3 to 1 now, where have other people started seeing problems at? _______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc _______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
Just was reviewing the "2 ARC" discussion on the list (I'm finally about to do this after all these years) and smiled at the reliability issue on the ARC's and NMC's. Just for grins I logged into one of my ARC's running 4.2.32 and lookie here: SYSTEM DESCRIPTION System Descriptor: 3Com Corporation HiPer Access Router Card Built on Aug 16 1999 at 17:19:37. Object ID: 1.3.6.1.4.1.429.2.19 System UpTime: 413d 07:07:16 System Version: V4.2.32 413 days! 4.2.32 has been by far the best code I've run on my ARC's, 5.2.102 is looking pretty good too, but still not enough time to say for sure. No memory leaks yet though, which is better than 5.2.9!! - Mike ------------------------------------------------ Rock Island Communications, Inc. (360)-378-5884 http://www.rockisland.com/ San Juan Islands, WA ------------------------------------------------
Hi. My MRTG graph (http://www.inet2000.com/mrtg/3comtc.modem.html) shows period with no modem use, however we've had no complaints of the system dying for any period of time and the call durations on some of our currently online users are as high as 47 hours for this call (Yes, that's a camper and another issue....) We're new to MRTG (as you can see from the graphs) and expect that this is a MRTG configuration issue (perhaps a longer timeout???) and not a real modem pool problem. Any help would be appreciated greatly. Todd Chamberlain support@inet2000.com
On Monday 26 August 2002 03:49 pm, Support wrote:
Hi. My MRTG graph (http://www.inet2000.com/mrtg/3comtc.modem.html) shows period with no modem use, however we've had no complaints of the system dying for any period of time and the call durations on some of our currently online users are as high as 47 hours for this call (Yes, that's a camper and another issue....) One of the options mrtg has is to reflect a null response as zero rather than the same as the last non-null response mrtg received. Maybe this is something to look into. More to the point would be to fix any network issues that are causing this timeout. -- Lewis Bergman Texas Communications 4309 Maple St. Abilene, TX 79602-8044 915-695-6962 ext 115
Thanks Lewis. I'm not certain it's a timeout at all - I was just guessing that there were actually people online during that time, just that MRTG didn't know it. I'm basically guessing that it didn't get a response quickly enough and records this as zero. I'm not aware of any network problems, but frankly that's using the 'everything seems fast and seems to work correctly' diagnostic tool. We've been running this modem pool for a couple month's and haven't had any substantial problems reported by users, and I'm pretty sure if we were offline for 45 minutes or so, 4 or 5 times yesterday, we'd have heard about it. :-) So, that means there really is no Modem Pool problem, just a MRTG problem - or at least some sort of reporting problem. The other reports on that same pool... http://www.inet2000.com/mrtg/3comtc.pub.html http://www.inet2000.com/mrtg/3comtc.ram.html http://www.inet2000.com/mrtg/modems.htm don't seem to have any gaps in the data, so it doesn't seem that MRTG has a hard time talking to it in other cases.... So, basically, I'm a newbie on MRTG and I don't really know what to ask, other than 'help.' :-) ----- Original Message ----- From: "Lewis Bergman" <lbergman@abi.tconline.net> To: <usr-tc@mailman.xmission.com> Sent: Monday, August 26, 2002 3:16 PM Subject: Re: [USR-TC] MRTG Help
On Monday 26 August 2002 03:49 pm, Support wrote:
Hi. My MRTG graph (http://www.inet2000.com/mrtg/3comtc.modem.html) shows period with no modem use, however we've had no complaints of the system dying for any period of time and the call durations on some of our currently online users are as high as 47 hours for this call (Yes, that's a camper and another issue....) One of the options mrtg has is to reflect a null response as zero rather than the same as the last non-null response mrtg received. Maybe this is something to look into. More to the point would be to fix any network issues that are causing this timeout. -- Lewis Bergman Texas Communications 4309 Maple St. Abilene, TX 79602-8044 915-695-6962 ext 115
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
I would try changing the MAX users setting to 26 or 28. It can be found in the MRTG config file. I have had problems where if the max threshhold is surpassed (could it be counting a telnet as a user?) my MRTG doesn't read it as zero and just repeats the last reading. (Using it for traffic not user count tho) There are other causes. Do you have a firewall between the MRTG box and the 3Com? What type of box are you running MRTG from? I have found that running MRTG from an NT box is flaky sometimes. This is my first post. yay! We bought a TC unit a while ago..not even sure what model it is..and I haven't set it up yet...hoping to get it done pretty soon...so I will have lots of stupid questions. I did notice that there is a lot of helpful stuff in the headers of the list email so I will try those links first.... Craig. -----Original Message----- From: usr-tc-admin@mailman.xmission.com [mailto:usr-tc-admin@mailman.xmission.com]On Behalf Of Support Sent: Monday, August 26, 2002 7:19 PM To: usr-tc@mailman.xmission.com Subject: Re: [USR-TC] MRTG Help Thanks Lewis. I'm not certain it's a timeout at all - I was just guessing that there were actually people online during that time, just that MRTG didn't know it. I'm basically guessing that it didn't get a response quickly enough and records this as zero. I'm not aware of any network problems, but frankly that's using the 'everything seems fast and seems to work correctly' diagnostic tool. We've been running this modem pool for a couple month's and haven't had any substantial problems reported by users, and I'm pretty sure if we were offline for 45 minutes or so, 4 or 5 times yesterday, we'd have heard about it. :-) So, that means there really is no Modem Pool problem, just a MRTG problem - or at least some sort of reporting problem. The other reports on that same pool... http://www.inet2000.com/mrtg/3comtc.pub.html http://www.inet2000.com/mrtg/3comtc.ram.html http://www.inet2000.com/mrtg/modems.htm don't seem to have any gaps in the data, so it doesn't seem that MRTG has a hard time talking to it in other cases.... So, basically, I'm a newbie on MRTG and I don't really know what to ask, other than 'help.' :-) ----- Original Message ----- From: "Lewis Bergman" <lbergman@abi.tconline.net> To: <usr-tc@mailman.xmission.com> Sent: Monday, August 26, 2002 3:16 PM Subject: Re: [USR-TC] MRTG Help
On Monday 26 August 2002 03:49 pm, Support wrote:
Hi. My MRTG graph (http://www.inet2000.com/mrtg/3comtc.modem.html) shows period with no modem use, however we've had no complaints of the system dying for any period of time and the call durations on some of our currently online users are as high as 47 hours for this call (Yes, that's a camper and another issue....) One of the options mrtg has is to reflect a null response as zero rather than the same as the last non-null response mrtg received. Maybe this is something to look into. More to the point would be to fix any network issues that are causing this timeout. -- Lewis Bergman Texas Communications 4309 Maple St. Abilene, TX 79602-8044 915-695-6962 ext 115
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
OK. Thanks Craig. I'm proud to be the recipient of your first post! I'll try these things. If you check the graph, it certainly does seem to be only a problem when that DSP card fills up, so it may well be the Max user thing. I'll check that, thanks. There is no firewall or anything, they're just sitting next to each other plugged into a 3COM PS40 Switching Hub. It is running on an NT machine, but we're running lots of Perl things running on NT machines. That's an idea, but I doubt it's the issue, since all the other MRTG graphs seem OK. Of course, I've only been MRTGing for a week and I have no clue. (those two statements are totally separate facts.) Todd Chamberlain support@inet2000.com ----- Original Message ----- From: "Craig Gittens" <cgittens@sunbeach.net> To: <usr-tc@mailman.xmission.com> Sent: Wednesday, August 28, 2002 9:10 AM Subject: RE: [USR-TC] MRTG Help
I would try changing the MAX users setting to 26 or 28. It can be found in the MRTG config file. I have had problems where if the max threshhold is surpassed (could it be counting a telnet as a user?) my MRTG doesn't read it as zero and just repeats the last reading. (Using it for traffic not user count tho) There are other causes. Do you have a firewall between the MRTG box and the 3Com? What type of box are you running MRTG from? I have found that running MRTG from an NT box is flaky sometimes.
This is my first post. yay! We bought a TC unit a while ago..not even sure what model it is..and I haven't set it up yet...hoping to get it done pretty soon...so I will have lots of stupid questions. I did notice that there is a lot of helpful stuff in the headers of the list email so I will try those links first....
Craig.
-----Original Message----- From: usr-tc-admin@mailman.xmission.com [mailto:usr-tc-admin@mailman.xmission.com]On Behalf Of Support Sent: Monday, August 26, 2002 7:19 PM To: usr-tc@mailman.xmission.com Subject: Re: [USR-TC] MRTG Help
Thanks Lewis. I'm not certain it's a timeout at all - I was just guessing that there were actually people online during that time, just that MRTG didn't know it. I'm basically guessing that it didn't get a response quickly enough and records this as zero.
I'm not aware of any network problems, but frankly that's using the 'everything seems fast and seems to work correctly' diagnostic tool. We've been running this modem pool for a couple month's and haven't had any substantial problems reported by users, and I'm pretty sure if we were offline for 45 minutes or so, 4 or 5 times yesterday, we'd have heard about it. :-)
So, that means there really is no Modem Pool problem, just a MRTG problem - or at least some sort of reporting problem. The other reports on that same pool...
http://www.inet2000.com/mrtg/3comtc.pub.html http://www.inet2000.com/mrtg/3comtc.ram.html http://www.inet2000.com/mrtg/modems.htm
don't seem to have any gaps in the data, so it doesn't seem that MRTG has a hard time talking to it in other cases....
So, basically, I'm a newbie on MRTG and I don't really know what to ask, other than 'help.' :-)
----- Original Message ----- From: "Lewis Bergman" <lbergman@abi.tconline.net> To: <usr-tc@mailman.xmission.com> Sent: Monday, August 26, 2002 3:16 PM Subject: Re: [USR-TC] MRTG Help
On Monday 26 August 2002 03:49 pm, Support wrote:
Hi. My MRTG graph (http://www.inet2000.com/mrtg/3comtc.modem.html) shows period with no modem use, however we've had no complaints of the system dying for any period of time and the call durations on some of our currently online users are as high as 47 hours for this call (Yes, that's a camper and another issue....) One of the options mrtg has is to reflect a null response as zero rather than the same as the last non-null response mrtg received. Maybe this is something to look into. More to the point would be to fix any network issues that are causing this timeout. -- Lewis Bergman Texas Communications 4309 Maple St. Abilene, TX 79602-8044 915-695-6962 ext 115
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
Actually, I think this is probably it - thank you. There is an option "unknaszero" (which I assume means to put any unknown response as a zero, and I should have a "unknown as last" instead. I'd rather not artificially put that max at 26, so do you happen to know what the correct option is for unknown as last value? Maybe 'unknaslast'? ----- Original Message ----- From: "Craig Gittens" <cgittens@sunbeach.net> To: <usr-tc@mailman.xmission.com> Sent: Wednesday, August 28, 2002 9:10 AM Subject: RE: [USR-TC] MRTG Help
I would try changing the MAX users setting to 26 or 28. It can be found in the MRTG config file. I have had problems where if the max threshhold is surpassed (could it be counting a telnet as a user?) my MRTG doesn't read it as zero and just repeats the last reading. (Using it for traffic not user count tho) There are other causes. Do you have a firewall between the MRTG box and the 3Com? What type of box are you running MRTG from? I have found that running MRTG from an NT box is flaky sometimes.
This is my first post. yay! We bought a TC unit a while ago..not even sure what model it is..and I haven't set it up yet...hoping to get it done pretty soon...so I will have lots of stupid questions. I did notice that there is a lot of helpful stuff in the headers of the list email so I will try those links first....
Craig.
-----Original Message----- From: usr-tc-admin@mailman.xmission.com [mailto:usr-tc-admin@mailman.xmission.com]On Behalf Of Support Sent: Monday, August 26, 2002 7:19 PM To: usr-tc@mailman.xmission.com Subject: Re: [USR-TC] MRTG Help
Thanks Lewis. I'm not certain it's a timeout at all - I was just guessing that there were actually people online during that time, just that MRTG didn't know it. I'm basically guessing that it didn't get a response quickly enough and records this as zero.
I'm not aware of any network problems, but frankly that's using the 'everything seems fast and seems to work correctly' diagnostic tool. We've been running this modem pool for a couple month's and haven't had any substantial problems reported by users, and I'm pretty sure if we were offline for 45 minutes or so, 4 or 5 times yesterday, we'd have heard about it. :-)
So, that means there really is no Modem Pool problem, just a MRTG problem - or at least some sort of reporting problem. The other reports on that same pool...
http://www.inet2000.com/mrtg/3comtc.pub.html http://www.inet2000.com/mrtg/3comtc.ram.html http://www.inet2000.com/mrtg/modems.htm
don't seem to have any gaps in the data, so it doesn't seem that MRTG has a hard time talking to it in other cases....
So, basically, I'm a newbie on MRTG and I don't really know what to ask, other than 'help.' :-)
----- Original Message ----- From: "Lewis Bergman" <lbergman@abi.tconline.net> To: <usr-tc@mailman.xmission.com> Sent: Monday, August 26, 2002 3:16 PM Subject: Re: [USR-TC] MRTG Help
On Monday 26 August 2002 03:49 pm, Support wrote:
Hi. My MRTG graph (http://www.inet2000.com/mrtg/3comtc.modem.html) shows period with no modem use, however we've had no complaints of the system dying for any period of time and the call durations on some of our currently online users are as high as 47 hours for this call (Yes, that's a camper and another issue....) One of the options mrtg has is to reflect a null response as zero rather than the same as the last non-null response mrtg received. Maybe this is something to look into. More to the point would be to fix any network issues that are causing this timeout. -- Lewis Bergman Texas Communications 4309 Maple St. Abilene, TX 79602-8044 915-695-6962 ext 115
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
I don't I am still working on MRTG 2.8 something. :) However putting it at 26 could be a temporary thing just to see if it is the problem. Craig. -----Original Message----- From: usr-tc-admin@mailman.xmission.com [mailto:usr-tc-admin@mailman.xmission.com]On Behalf Of Support Sent: Wednesday, August 28, 2002 12:56 PM To: usr-tc@mailman.xmission.com Subject: Re: [USR-TC] MRTG Help Actually, I think this is probably it - thank you. There is an option "unknaszero" (which I assume means to put any unknown response as a zero, and I should have a "unknown as last" instead. I'd rather not artificially put that max at 26, so do you happen to know what the correct option is for unknown as last value? Maybe 'unknaslast'? ----- Original Message ----- From: "Craig Gittens" <cgittens@sunbeach.net> To: <usr-tc@mailman.xmission.com> Sent: Wednesday, August 28, 2002 9:10 AM Subject: RE: [USR-TC] MRTG Help
I would try changing the MAX users setting to 26 or 28. It can be found in the MRTG config file. I have had problems where if the max threshhold is surpassed (could it be counting a telnet as a user?) my MRTG doesn't read it as zero and just repeats the last reading. (Using it for traffic not user count tho) There are other causes. Do you have a firewall between the MRTG box and the 3Com? What type of box are you running MRTG from? I have found that running MRTG from an NT box is flaky sometimes.
This is my first post. yay! We bought a TC unit a while ago..not even sure what model it is..and I haven't set it up yet...hoping to get it done pretty soon...so I will have lots of stupid questions. I did notice that there is a lot of helpful stuff in the headers of the list email so I will try those links first....
Craig.
-----Original Message----- From: usr-tc-admin@mailman.xmission.com [mailto:usr-tc-admin@mailman.xmission.com]On Behalf Of Support Sent: Monday, August 26, 2002 7:19 PM To: usr-tc@mailman.xmission.com Subject: Re: [USR-TC] MRTG Help
Thanks Lewis. I'm not certain it's a timeout at all - I was just guessing that there were actually people online during that time, just that MRTG didn't know it. I'm basically guessing that it didn't get a response quickly enough and records this as zero.
I'm not aware of any network problems, but frankly that's using the 'everything seems fast and seems to work correctly' diagnostic tool. We've been running this modem pool for a couple month's and haven't had any substantial problems reported by users, and I'm pretty sure if we were offline for 45 minutes or so, 4 or 5 times yesterday, we'd have heard about it. :-)
So, that means there really is no Modem Pool problem, just a MRTG problem - or at least some sort of reporting problem. The other reports on that same pool...
http://www.inet2000.com/mrtg/3comtc.pub.html http://www.inet2000.com/mrtg/3comtc.ram.html http://www.inet2000.com/mrtg/modems.htm
don't seem to have any gaps in the data, so it doesn't seem that MRTG has a hard time talking to it in other cases....
So, basically, I'm a newbie on MRTG and I don't really know what to ask, other than 'help.' :-)
----- Original Message ----- From: "Lewis Bergman" <lbergman@abi.tconline.net> To: <usr-tc@mailman.xmission.com> Sent: Monday, August 26, 2002 3:16 PM Subject: Re: [USR-TC] MRTG Help
On Monday 26 August 2002 03:49 pm, Support wrote:
Hi. My MRTG graph (http://www.inet2000.com/mrtg/3comtc.modem.html) shows period with no modem use, however we've had no complaints of the system dying for any period of time and the call durations on some of our currently online users are as high as 47 hours for this call (Yes, that's a camper and another issue....) One of the options mrtg has is to reflect a null response as zero rather than the same as the last non-null response mrtg received. Maybe this is something to look into. More to the point would be to fix any network issues that are causing this timeout. -- Lewis Bergman Texas Communications 4309 Maple St. Abilene, TX 79602-8044 915-695-6962 ext 115
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
Craig, thank you. This turned out to be the problem - at least so far it looks good. I read the MRTG manual (NO!) and the 'unknaszero' option is exactly that, an option. If you leave it off, it'll automatically do a 'unknown as last value' sort of situation. I simply removed that option and now I left the MAX at 24 (which reflect the modems in that pool) and we look set. You can see the difference from last evening to so far this evening with no dropout to zero's so far http://www.inet2000.com/mrtg/3comtc.modem.html Thanks again, Craig. Todd Chamberlain ----- Original Message ----- From: "Craig Gittens" <cgittens@sunbeach.net> To: <usr-tc@mailman.xmission.com> Sent: Wednesday, August 28, 2002 11:23 AM Subject: RE: [USR-TC] MRTG Help
I don't I am still working on MRTG 2.8 something. :) However putting it at 26 could be a temporary thing just to see if it is the problem.
Craig.
-----Original Message----- From: usr-tc-admin@mailman.xmission.com [mailto:usr-tc-admin@mailman.xmission.com]On Behalf Of Support Sent: Wednesday, August 28, 2002 12:56 PM To: usr-tc@mailman.xmission.com Subject: Re: [USR-TC] MRTG Help
Actually, I think this is probably it - thank you. There is an option "unknaszero" (which I assume means to put any unknown response as a zero, and I should have a "unknown as last" instead. I'd rather not artificially put that max at 26, so do you happen to know what the correct option is for unknown as last value? Maybe 'unknaslast'?
----- Original Message ----- From: "Craig Gittens" <cgittens@sunbeach.net> To: <usr-tc@mailman.xmission.com> Sent: Wednesday, August 28, 2002 9:10 AM Subject: RE: [USR-TC] MRTG Help
I would try changing the MAX users setting to 26 or 28. It can be found in the MRTG config file. I have had problems where if the max threshhold is surpassed (could it be counting a telnet as a user?) my MRTG doesn't read it as zero and just repeats the last reading. (Using it for traffic not user count tho) There are other causes. Do you have a firewall between the MRTG box and the 3Com? What type of box are you running MRTG from? I have found that running MRTG from an NT box is flaky sometimes.
This is my first post. yay! We bought a TC unit a while ago..not even sure what model it is..and I haven't set it up yet...hoping to get it done pretty soon...so I will have lots of stupid questions. I did notice that there is a lot of helpful stuff in the headers of the list email so I will try those links first....
Craig.
-----Original Message----- From: usr-tc-admin@mailman.xmission.com [mailto:usr-tc-admin@mailman.xmission.com]On Behalf Of Support Sent: Monday, August 26, 2002 7:19 PM To: usr-tc@mailman.xmission.com Subject: Re: [USR-TC] MRTG Help
Thanks Lewis. I'm not certain it's a timeout at all - I was just guessing that there were actually people online during that time, just that MRTG didn't know it. I'm basically guessing that it didn't get a response quickly enough and records this as zero.
I'm not aware of any network problems, but frankly that's using the 'everything seems fast and seems to work correctly' diagnostic tool. We've been running this modem pool for a couple month's and haven't had any substantial problems reported by users, and I'm pretty sure if we were offline for 45 minutes or so, 4 or 5 times yesterday, we'd have heard about it. :-)
So, that means there really is no Modem Pool problem, just a MRTG problem - or at least some sort of reporting problem. The other reports on that same pool...
http://www.inet2000.com/mrtg/3comtc.pub.html http://www.inet2000.com/mrtg/3comtc.ram.html http://www.inet2000.com/mrtg/modems.htm
don't seem to have any gaps in the data, so it doesn't seem that MRTG has a hard time talking to it in other cases....
So, basically, I'm a newbie on MRTG and I don't really know what to ask, other than 'help.' :-)
----- Original Message ----- From: "Lewis Bergman" <lbergman@abi.tconline.net> To: <usr-tc@mailman.xmission.com> Sent: Monday, August 26, 2002 3:16 PM Subject: Re: [USR-TC] MRTG Help
On Monday 26 August 2002 03:49 pm, Support wrote:
Hi. My MRTG graph (http://www.inet2000.com/mrtg/3comtc.modem.html) shows period with no modem use, however we've had no complaints of the system dying for any period of time and the call durations on some of our currently online users are as high as 47 hours for this call (Yes, that's a camper and another issue....) One of the options mrtg has is to reflect a null response as zero rather than the same as the last non-null response mrtg received. Maybe this is something to look into. More to the point would be to fix any network issues that are causing this timeout. -- Lewis Bergman Texas Communications 4309 Maple St. Abilene, TX 79602-8044 915-695-6962 ext 115
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
I'm glad that fixed it Todd. However I must say that I would not be happy with that solution. Why is it possibly getting a reading over 24 or an unknown reading? That would bug the h311 out of me. For example, suppose you get a reading of 6 and then get that flat 6 reading for the next two hours because it suddenly jumped to 25 or some bug kicked in. You have no reliable stats. Yes, your graphs looks prettier now but not necessarily correct. If you notice you have quite a few flat lines in it which indicate it did it's unknown number for at least 15-20 minutes. That is an indication that something is wrong. Craig. -----Original Message----- From: usr-tc-admin@mailman.xmission.com [mailto:usr-tc-admin@mailman.xmission.com]On Behalf Of Support Sent: Wednesday, August 28, 2002 8:23 PM To: usr-tc@mailman.xmission.com Subject: Re: [USR-TC] MRTG Help Craig, thank you. This turned out to be the problem - at least so far it looks good. I read the MRTG manual (NO!) and the 'unknaszero' option is exactly that, an option. If you leave it off, it'll automatically do a 'unknown as last value' sort of situation. I simply removed that option and now I left the MAX at 24 (which reflect the modems in that pool) and we look set. You can see the difference from last evening to so far this evening with no dropout to zero's so far http://www.inet2000.com/mrtg/3comtc.modem.html Thanks again, Craig. Todd Chamberlain ----- Original Message ----- From: "Craig Gittens" <cgittens@sunbeach.net> To: <usr-tc@mailman.xmission.com> Sent: Wednesday, August 28, 2002 11:23 AM Subject: RE: [USR-TC] MRTG Help
I don't I am still working on MRTG 2.8 something. :) However putting it at 26 could be a temporary thing just to see if it is the problem.
Craig.
-----Original Message----- From: usr-tc-admin@mailman.xmission.com [mailto:usr-tc-admin@mailman.xmission.com]On Behalf Of Support Sent: Wednesday, August 28, 2002 12:56 PM To: usr-tc@mailman.xmission.com Subject: Re: [USR-TC] MRTG Help
Actually, I think this is probably it - thank you. There is an option "unknaszero" (which I assume means to put any unknown response as a zero, and I should have a "unknown as last" instead. I'd rather not artificially put that max at 26, so do you happen to know what the correct option is for unknown as last value? Maybe 'unknaslast'?
----- Original Message ----- From: "Craig Gittens" <cgittens@sunbeach.net> To: <usr-tc@mailman.xmission.com> Sent: Wednesday, August 28, 2002 9:10 AM Subject: RE: [USR-TC] MRTG Help
I would try changing the MAX users setting to 26 or 28. It can be found in the MRTG config file. I have had problems where if the max threshhold is surpassed (could it be counting a telnet as a user?) my MRTG doesn't read it as zero and just repeats the last reading. (Using it for traffic not user count tho) There are other causes. Do you have a firewall between the MRTG box and the 3Com? What type of box are you running MRTG from? I have found that running MRTG from an NT box is flaky sometimes.
This is my first post. yay! We bought a TC unit a while ago..not even sure what model it is..and I haven't set it up yet...hoping to get it done pretty soon...so I will have lots of stupid questions. I did notice that there is a lot of helpful stuff in the headers of the list email so I will try those links first....
Craig.
-----Original Message----- From: usr-tc-admin@mailman.xmission.com [mailto:usr-tc-admin@mailman.xmission.com]On Behalf Of Support Sent: Monday, August 26, 2002 7:19 PM To: usr-tc@mailman.xmission.com Subject: Re: [USR-TC] MRTG Help
Thanks Lewis. I'm not certain it's a timeout at all - I was just guessing that there were actually people online during that time, just that MRTG didn't know it. I'm basically guessing that it didn't get a response quickly enough and records this as zero.
I'm not aware of any network problems, but frankly that's using the 'everything seems fast and seems to work correctly' diagnostic tool. We've been running this modem pool for a couple month's and haven't had any substantial problems reported by users, and I'm pretty sure if we were offline for 45 minutes or so, 4 or 5 times yesterday, we'd have heard about it. :-)
So, that means there really is no Modem Pool problem, just a MRTG problem - or at least some sort of reporting problem. The other reports on that same pool...
http://www.inet2000.com/mrtg/3comtc.pub.html http://www.inet2000.com/mrtg/3comtc.ram.html http://www.inet2000.com/mrtg/modems.htm
don't seem to have any gaps in the data, so it doesn't seem that MRTG has a hard time talking to it in other cases....
So, basically, I'm a newbie on MRTG and I don't really know what to ask, other than 'help.' :-)
----- Original Message ----- From: "Lewis Bergman" <lbergman@abi.tconline.net> To: <usr-tc@mailman.xmission.com> Sent: Monday, August 26, 2002 3:16 PM Subject: Re: [USR-TC] MRTG Help
On Monday 26 August 2002 03:49 pm, Support wrote:
Hi. My MRTG graph (http://www.inet2000.com/mrtg/3comtc.modem.html) shows period with no modem use, however we've had no complaints of the system dying for any period of time and the call durations on some of our currently online users are as high as 47 hours for this call (Yes, that's a camper and another issue....) One of the options mrtg has is to reflect a null response as zero rather than the same as the last non-null response mrtg received. Maybe this is something to look into. More to the point would be to fix any network issues that are causing this timeout. -- Lewis Bergman Texas Communications 4309 Maple St. Abilene, TX 79602-8044 915-695-6962 ext 115
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
I agree, sort of... Because all my other graphs seem OK (at least there is no compelling - glaring reason to think they are wrong) my gut reaction is that the MRTG is working OK and communicating with the 3Com pool Ok and everthing, and it's just adding one user for a telnet session, or it's seeing the last user not released yet while the next user is already connected or something. I had assumed that the Pool was reporting a telnet session or something as an extra active connection, so that when the card was full with 24 real people online, it was reporting 25 to MRTG and MRTG saw that as an unknown value. Maybe this is a bad assumption. In any case, a guy shouldn't assume things. (you know what they say about AssUMeing things.) I had also thought that setting Max Modems to 26 or 28 wasn't a good idea since that would put a 26 or 28 along the left side of the graph, and would throw all the Percentages out, but that may be the better thing, at least to test. This way we can see if the peak modems in use are 25 or 26 or 27 and see how far out the reported numbers are. If my average users are 16, I want to know that it IS actually 16 and not really 14 plus two telnet sessions. Thanks, I've done that and we'll wait a while and see how it looks. :-) http://www.inet2000.com/mrtg/3comtc.modem.html Todd Chamberlain ----- Original Message ----- From: "Craig Gittens" <cgittens@sunbeach.net> To: <usr-tc@mailman.xmission.com> Sent: Thursday, August 29, 2002 6:30 AM Subject: RE: [USR-TC] MRTG Help
I'm glad that fixed it Todd. However I must say that I would not be happy with that solution. Why is it possibly getting a reading over 24 or an unknown reading? That would bug the h311 out of me. For example, suppose you get a reading of 6 and then get that flat 6 reading for the next two hours because it suddenly jumped to 25 or some bug kicked in. You have no reliable stats. Yes, your graphs looks prettier now but not necessarily correct. If you notice you have quite a few flat lines in it which indicate it did it's unknown number for at least 15-20 minutes. That is an indication that something is wrong.
Craig.
-----Original Message----- From: usr-tc-admin@mailman.xmission.com [mailto:usr-tc-admin@mailman.xmission.com]On Behalf Of Support Sent: Wednesday, August 28, 2002 8:23 PM To: usr-tc@mailman.xmission.com Subject: Re: [USR-TC] MRTG Help
Craig, thank you. This turned out to be the problem - at least so far it looks good. I read the MRTG manual (NO!) and the 'unknaszero' option is exactly that, an option. If you leave it off, it'll automatically do a 'unknown as last value' sort of situation. I simply removed that option and now I left the MAX at 24 (which reflect the modems in that pool) and we look set. You can see the difference from last evening to so far this evening with no dropout to zero's so far
http://www.inet2000.com/mrtg/3comtc.modem.html
Thanks again, Craig.
Todd Chamberlain
----- Original Message ----- From: "Craig Gittens" <cgittens@sunbeach.net> To: <usr-tc@mailman.xmission.com> Sent: Wednesday, August 28, 2002 11:23 AM Subject: RE: [USR-TC] MRTG Help
I don't I am still working on MRTG 2.8 something. :) However putting it at 26 could be a temporary thing just to see if it is the problem.
Craig.
-----Original Message----- From: usr-tc-admin@mailman.xmission.com [mailto:usr-tc-admin@mailman.xmission.com]On Behalf Of Support Sent: Wednesday, August 28, 2002 12:56 PM To: usr-tc@mailman.xmission.com Subject: Re: [USR-TC] MRTG Help
Actually, I think this is probably it - thank you. There is an option "unknaszero" (which I assume means to put any unknown response as a zero, and I should have a "unknown as last" instead. I'd rather not artificially put that max at 26, so do you happen to know what the correct option is for unknown as last value? Maybe 'unknaslast'?
----- Original Message ----- From: "Craig Gittens" <cgittens@sunbeach.net> To: <usr-tc@mailman.xmission.com> Sent: Wednesday, August 28, 2002 9:10 AM Subject: RE: [USR-TC] MRTG Help
I would try changing the MAX users setting to 26 or 28. It can be found in the MRTG config file. I have had problems where if the max threshhold is surpassed (could it be counting a telnet as a user?) my MRTG doesn't read it as zero and just repeats the last reading. (Using it for traffic not user count tho) There are other causes. Do you have a firewall between the MRTG box and the 3Com? What type of box are you running MRTG from? I have found that running MRTG from an NT box is flaky sometimes.
This is my first post. yay! We bought a TC unit a while ago..not even sure what model it is..and I haven't set it up yet...hoping to get it done pretty soon...so I will have lots of stupid questions. I did notice that there is a lot of helpful stuff in the headers of the list email so I will try those links first....
Craig.
-----Original Message----- From: usr-tc-admin@mailman.xmission.com [mailto:usr-tc-admin@mailman.xmission.com]On Behalf Of Support Sent: Monday, August 26, 2002 7:19 PM To: usr-tc@mailman.xmission.com Subject: Re: [USR-TC] MRTG Help
Thanks Lewis. I'm not certain it's a timeout at all - I was just guessing that there were actually people online during that time, just that MRTG didn't know it. I'm basically guessing that it didn't get a response quickly enough and records this as zero.
I'm not aware of any network problems, but frankly that's using the 'everything seems fast and seems to work correctly' diagnostic tool. We've been running this modem pool for a couple month's and haven't had any substantial problems reported by users, and I'm pretty sure if we were offline for 45 minutes or so, 4 or 5 times yesterday, we'd have heard about it. :-)
So, that means there really is no Modem Pool problem, just a MRTG problem - or at least some sort of reporting problem. The other reports on that same pool...
http://www.inet2000.com/mrtg/3comtc.pub.html http://www.inet2000.com/mrtg/3comtc.ram.html http://www.inet2000.com/mrtg/modems.htm
don't seem to have any gaps in the data, so it doesn't seem that MRTG has a hard time talking to it in other cases....
So, basically, I'm a newbie on MRTG and I don't really know what to ask, other than 'help.' :-)
----- Original Message ----- From: "Lewis Bergman" <lbergman@abi.tconline.net> To: <usr-tc@mailman.xmission.com> Sent: Monday, August 26, 2002 3:16 PM Subject: Re: [USR-TC] MRTG Help
On Monday 26 August 2002 03:49 pm, Support wrote:
Hi. My MRTG graph (http://www.inet2000.com/mrtg/3comtc.modem.html) shows period with no modem use, however we've had no complaints of the system dying for any period of time and the call durations on some of our currently online users are as high as 47 hours for this call (Yes, that's a camper and another issue....) One of the options mrtg has is to reflect a null response as zero rather than the same as the last non-null response mrtg received. Maybe this is something to look into. More to the point would be to fix any network issues that are causing this timeout. -- Lewis Bergman Texas Communications 4309 Maple St. Abilene, TX 79602-8044 915-695-6962 ext 115
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
On Thursday 29 August 2002 07:30 am, Craig Gittens wrote:
I'm glad that fixed it Todd. However I must say that I would not be happy with that solution. Why is it possibly getting a reading over 24 or an unknown reading? That would bug the h311 out of me. For example, suppose you get a reading of 6 and then get that flat 6 reading for the next two hours because it suddenly jumped to 25 or some bug kicked in. You have no reliable stats. Yes, your graphs looks prettier now but not necessarily correct. If you notice you have quite a few flat lines in it which indicate it did it's unknown number for at least 15-20 minutes. That is an indication that something is wrong. The real solution of course is to buy enough ports so that one or two telnet users doesn't mean your graphs are totally whacked. These things are chep enough now, just buy another. -- Lewis Bergman Texas Communications 4309 Maple St. Abilene, TX 79602-8044 915-695-6962 ext 115
Hey, do you have a deal with my Telco on C-T1 lines too? :-) The HiperDSP cards are cheap, but the Telco pricing isn't. Actually, I have two HDSP cards, but only have one C-T1 right now. Some of the pricing here is based on escalation the more you have, so we have to combine some ports on T1, some on BRI and some on Analog Centrex lines. For example, if you call up the telco and want 1 BRI line (so two telephone or modem channels) the BRI line will be $74.50 each (so $37.25 per line). As you would normally expect, prices get a bit cheaper for 2 lines or 5 lines, so that they eventually get down to $62.50 per BRI (just $31.25 per modem line) for 10 BRI lines. However, if you want 11+ lines, suddenly ALL the lines (including the first 10) go up to $150 per line. SaskTel says this is due to trunk loading (Yea, that's a load alright) but it's obviously to keep competative ISP's in the red. 10 lines cost $500 / month, but 11 lines would cost $1,650 / month. That's a pretty expensive 11th line! The same thing with Centrex lines - they start off affordable, but if you get to many in the same hunt group, the price escalates to make it unaffordable. So, we have 10 BRI lines (20 modems), 16 Centrex lines and a T1 (24 modem ports) for 60 ports. ----- Original Message ----- From: "Lewis Bergman" <lbergman@abi.tconline.net> To: <usr-tc@mailman.xmission.com> Sent: Thursday, August 29, 2002 2:54 PM Subject: Re: [USR-TC] MRTG Help
On Thursday 29 August 2002 07:30 am, Craig Gittens wrote:
I'm glad that fixed it Todd. However I must say that I would not be happy with that solution. Why is it possibly getting a reading over 24 or an unknown reading? That would bug the h311 out of me. For example, suppose you get a reading of 6 and then get that flat 6 reading for the next two hours because it suddenly jumped to 25 or some bug kicked in. You have no reliable stats. Yes, your graphs looks prettier now but not necessarily correct. If you notice you have quite a few flat lines in it which indicate it did it's unknown number for at least 15-20 minutes. That is an indication that something is wrong. The real solution of course is to buy enough ports so that one or two telnet users doesn't mean your graphs are totally whacked. These things are chep enough now, just buy another. -- Lewis Bergman Texas Communications 4309 Maple St. Abilene, TX 79602-8044 915-695-6962 ext 115
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
And I thought we had it bad with South Western Bell. They are anti compettitive and illegally bundeling regulated and non-regulated services but they aren't that blatent. What country are you in?
Guess. The government is totally charge and squashes any innovation by the private individual. Russia of course. :-) No sorry, actually Saskatchewan, Canada. SaskTel is our Government run Telco. Todd Chamberlain ----- Original Message ----- From: "Lewis Bergman" <lbergman@abi.tconline.net> To: <usr-tc@mailman.xmission.com> Sent: Friday, August 30, 2002 6:04 PM Subject: Re: [USR-TC] MRTG Help
And I thought we had it bad with South Western Bell. They are anti compettitive and illegally bundeling regulated and non-regulated services but they aren't that blatent.
What country are you in?
_______________________________________________ USR-TC mailing list USR-TC@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/usr-tc
participants (10)
-
Aaron Nabil -
Chris A. Kalin -
Craig Gittens -
Lewis Bergman -
MacDonald, James -
Mike Greene -
Support -
Todd Bertolozzi -
Tom Williams -
V