Re: [Asterisk-Users] Digium Quad Span Cards
On Wed, 27 Apr 2005, Adam Goryachev wrote: Just wondering, but does the AMD multi CPU architecture improve the interrupt handling? My understanding of that architecture is that each CPU can deal with it's own PCI bus/interrupts/etc independently of each other, and also with their own memory/etc? Would this improve the scalability? In fact, would a multi-PCI bus system by itself 'solve' the problem? Beware that not all multi cpu Opteron motherboards are created equal. Quite a few connect all their pci busses to one cpu. A good motherboard will distribute the pci busses across the cpus. Read http://www.samag.com/documents/s=9408/sam0411b/0411b.htm for good and bad examples and a list of things to watch out for when purchasing an Opteron system. Peter ___ Asterisk-Users mailing list Asterisk-Users@lists.digium.com http://lists.digium.com/mailman/listinfo/asterisk-users To UNSUBSCRIBE or update options visit: http://lists.digium.com/mailman/listinfo/asterisk-users
Re: [Asterisk-Users] Digium Quad Span Cards
Initially, I believed that the limitation was the PCI bus, but I was mistaken. There is a lot of confusion surrounding this issue, and it would be great if someone stepped forward with a concrete answer. That said, here's what I've learned about the issue through my research. We started off asking a Digium representative about putting 4 quad-span cards in a single machine and got the following response: - I'd use two machines, with two quad cards each. And then, I'd need to be using only the G.711 ulaw protocol. Then, I'd still use a mid-range dual Xeon CPU machine in the 2.4GHz+ spectrum. David Mandelstam of the Asterisk Biz List provided me with this explanation: - Zaptel drivers produce at least one interrupt per millisecond per board, which is minimal if the interrupt handlers are short. But on a heavily loaded machine doing lots of echo cancellation, each interrupt can approach 1 millisecond in length. So if there are interrupts coming from several cards, you can see how you could get into trouble. So it looks like processor interrupts are the culprit. Possible solutions to this problem include (please feel free to add to this list): - An Asterisk slave server pool ( http://home.comcast.net/~mroth01/LargeAsteriskSetup.gif ) - A TDM-VoIP gateway (Cisco, Quintum, AudioCodes, Lucent) - Using Sangoma cards (As per David Mandelstam, Sangoma cards use proprietary drivers and there are operational setups using 4 quads per machine) I'm not confident that the Asterisk software scales well under certain conditions, such as using Monitor to digitally record 16 spans of voice channels, so solving the card issue may not be the last step in a large installation. If anyone has any insight on this, please post it to the list. Hope that was helpful, Matthew Roth http://www.voip-info.org/tiki-index.php?page=Running%20Asterisk%20on%20Debian Matthew Boehm wrote: From what I understand (and this could be completely wrong), the Digium cards use a bunch of processor interrupts and too many cards will use up all the interrupts. (again, that could be completely wrong). What kind of calls are they? G711-PRI? Not much CPU needed there. G729 - PRI? Yes, you would need something along the lines of a dual Xenon 3.6Ghz do do that. Its all in the transcoding. If just passthru, not much cpu is needed. -Matthew ___ Asterisk-Users mailing list Asterisk-Users@lists.digium.com http://lists.digium.com/mailman/listinfo/asterisk-users To UNSUBSCRIBE or update options visit: http://lists.digium.com/mailman/listinfo/asterisk-users
RE: [Asterisk-Users] Digium Quad Span Cards
That seams to be the same issue with SpanDSP. It seams that the high interrupt rate is slipping. In the case of the SpanDSP issue it is drop 1 out of 50 packets. This is of course with the TDM cards (fxo/fxs) not the Single or Quad span cards. I think it may be time to look at the Zap vased code to see if buffering or interrupt queues of a sort may be needed. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Matt Roth Sent: Tuesday, April 26, 2005 12:58 PM To: Asterisk Users Mailing List - Non-Commercial Discussion Subject: Re: [Asterisk-Users] Digium Quad Span Cards Initially, I believed that the limitation was the PCI bus, but I was mistaken. There is a lot of confusion surrounding this issue, and it would be great if someone stepped forward with a concrete answer. That said, here's what I've learned about the issue through my research. We started off asking a Digium representative about putting 4 quad-span cards in a single machine and got the following response: - I'd use two machines, with two quad cards each. And then, I'd need to be using only the G.711 ulaw protocol. Then, I'd still use a mid-range dual Xeon CPU machine in the 2.4GHz+ spectrum. David Mandelstam of the Asterisk Biz List provided me with this explanation: - Zaptel drivers produce at least one interrupt per millisecond per board, which is minimal if the interrupt handlers are short. But on a heavily loaded machine doing lots of echo cancellation, each interrupt can approach 1 millisecond in length. So if there are interrupts coming from several cards, you can see how you could get into trouble. So it looks like processor interrupts are the culprit. Possible solutions to this problem include (please feel free to add to this list): - An Asterisk slave server pool ( http://home.comcast.net/~mroth01/LargeAsteriskSetup.gif ) - A TDM-VoIP gateway (Cisco, Quintum, AudioCodes, Lucent) - Using Sangoma cards (As per David Mandelstam, Sangoma cards use proprietary drivers and there are operational setups using 4 quads per machine) I'm not confident that the Asterisk software scales well under certain conditions, such as using Monitor to digitally record 16 spans of voice channels, so solving the card issue may not be the last step in a large installation. If anyone has any insight on this, please post it to the list. Hope that was helpful, Matthew Roth http://www.voip-info.org/tiki-index.php?page=Running%20Asterisk%20on%20D ebian Matthew Boehm wrote: From what I understand (and this could be completely wrong), the Digium cards use a bunch of processor interrupts and too many cards will use up all the interrupts. (again, that could be completely wrong). What kind of calls are they? G711-PRI? Not much CPU needed there. G729 - PRI? Yes, you would need something along the lines of a dual Xenon 3.6Ghz do do that. Its all in the transcoding. If just passthru, not much cpu is needed. -Matthew ___ Asterisk-Users mailing list Asterisk-Users@lists.digium.com http://lists.digium.com/mailman/listinfo/asterisk-users To UNSUBSCRIBE or update options visit: http://lists.digium.com/mailman/listinfo/asterisk-users ___ Asterisk-Users mailing list Asterisk-Users@lists.digium.com http://lists.digium.com/mailman/listinfo/asterisk-users To UNSUBSCRIBE or update options visit: http://lists.digium.com/mailman/listinfo/asterisk-users
Re: [Asterisk-Users] Digium Quad Span Cards
On Tue, 2005-04-26 at 12:58 -0400, Matt Roth wrote: So it looks like processor interrupts are the culprit. Possible solutions to this problem include (please feel free to add to this list): - An Asterisk slave server pool ( http://home.comcast.net/~mroth01/LargeAsteriskSetup.gif ) - A TDM-VoIP gateway (Cisco, Quintum, AudioCodes, Lucent) - Using Sangoma cards (As per David Mandelstam, Sangoma cards use proprietary drivers and there are operational setups using 4 quads per machine) Just wondering, but does the AMD multi CPU architecture improve the interrupt handling? My understanding of that architecture is that each CPU can deal with it's own PCI bus/interrupts/etc independently of each other, and also with their own memory/etc? Would this improve the scalability? In fact, would a multi-PCI bus system by itself 'solve' the problem? I'm not confident that the Asterisk software scales well under certain conditions, such as using Monitor to digitally record 16 spans of voice channels, so solving the card issue may not be the last step in a large installation. If anyone has any insight on this, please post it to the list. Well, what are the overheads of monitoring a call as opposed to simply bridging it from a digium channel to a IP channel (ie, the voice is still passing through asterisk)? As I see it, you have: * Conversion from ulaw/alaw - slinear * Conversion from slinear - file format (what if you record in alaw/ulaw?) * Disk subsystem (writing to the disk) On a suitable system, I think the CPU involved in the transcoding from ulaw/alaw to slinear would be minimal. Converting from slinear to gsm for example, might be quite high, but if you record in ulaw/alaw, then this might work. Also, CPU overhead on disk performance should be minimal if using a reasonable SCSI controller. Lets see, 64000 /8 * 30 * 8 = 192 bytes/sec bps bits per bytechannels spans So, we are only writing 2 Mega Bytes / sec to the disk. That isn't exactly a lot of load for a disk to handle Sure, sustained transfer rates, interrupts, etc but really, 2MB/sec seems so slow :) So, is it really an issue? Dunno, someone want to run a couple of spans through monitor and try it out? We won't really know until we try it... Just my 2c worth... -- -- Adam Goryachev Website Managers Ph: +61 2 8304 [EMAIL PROTECTED] Fax: +61 2 9345 4396www.websitemanagers.com.au ___ Asterisk-Users mailing list Asterisk-Users@lists.digium.com http://lists.digium.com/mailman/listinfo/asterisk-users To UNSUBSCRIBE or update options visit: http://lists.digium.com/mailman/listinfo/asterisk-users
Re: [Asterisk-Users] Digium Quad Span Cards
On Apr 26, 2005, at 9:36 PM, Adam Goryachev wrote: Just wondering, but does the AMD multi CPU architecture improve the interrupt handling? My understanding of that architecture is that each CPU can deal with it's own PCI bus/interrupts/etc independently of each other, and also with their own memory/etc? Would this improve the scalability? In fact, would a multi-PCI bus system by itself 'solve' the problem? I'd be interested in knowing this as well. I'm not confident that the Asterisk software scales well under certain conditions, such as using Monitor to digitally record 16 spans of voice channels, so solving the card issue may not be the last step in a large installation. If anyone has any insight on this, please post it to the list. Can't prove it, but by reading this: http://www.sineapps.com/news.php?rssid=653 it makes reference to a heavy loaded asterisk box doing a lot of monitoring. On a suitable system, I think the CPU involved in the transcoding from ulaw/alaw to slinear would be minimal. Converting from slinear to gsm for example, might be quite high, but if you record in ulaw/alaw, then this might work. How do you specify how you want Monitor to save the audio. Sorry for my ignorance. Thanks, Daniel ___ Asterisk-Users mailing list Asterisk-Users@lists.digium.com http://lists.digium.com/mailman/listinfo/asterisk-users To UNSUBSCRIBE or update options visit: http://lists.digium.com/mailman/listinfo/asterisk-users
[Asterisk-Users] Digium Quad Span Cards
Hi all, Just a quick question for everyone. I understand that there is problems using more than 1 digium quad card in a single machine. Can anyone explain to me what the problems are? Also, I was wondering what the minimum spec would be for a * machine with a single quad card that was running a total of 120 concurrent calls. Can someone give me some real world examples of what they might have used in the field ? Thanks, Callum ___ Asterisk-Users mailing list Asterisk-Users@lists.digium.com http://lists.digium.com/mailman/listinfo/asterisk-users To UNSUBSCRIBE or update options visit: http://lists.digium.com/mailman/listinfo/asterisk-users
Re: [Asterisk-Users] Digium Quad Span Cards
From what I understand (and this could be completely wrong), the Digium cards use a bunch of processor interrupts and too many cards will use up all the interrupts. (again, that could be completely wrong). What kind of calls are they? G711-PRI? Not much CPU needed there. G729 - PRI? Yes, you would need something along the lines of a dual Xenon 3.6Ghz do do that. Its all in the transcoding. If just passthru, not much cpu is needed. -Matthew From: Callum McGillivray [EMAIL PROTECTED] Reply-To: Asterisk Users Mailing List - Non-Commercial Discussion asterisk-users@lists.digium.com Date: Tue, 26 Apr 2005 13:50:54 +1000 To: asterisk-users@lists.digium.com Subject: [Asterisk-Users] Digium Quad Span Cards Hi all, Just a quick question for everyone. I understand that there is problems using more than 1 digium quad card in a single machine. Can anyone explain to me what the problems are? Also, I was wondering what the minimum spec would be for a * machine with a single quad card that was running a total of 120 concurrent calls. Can someone give me some real world examples of what they might have used in the field ? Thanks, Callum ___ Asterisk-Users mailing list Asterisk-Users@lists.digium.com http://lists.digium.com/mailman/listinfo/asterisk-users To UNSUBSCRIBE or update options visit: http://lists.digium.com/mailman/listinfo/asterisk-users ___ Asterisk-Users mailing list Asterisk-Users@lists.digium.com http://lists.digium.com/mailman/listinfo/asterisk-users To UNSUBSCRIBE or update options visit: http://lists.digium.com/mailman/listinfo/asterisk-users
