-
1
- #1
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 strongm on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!
Voice over IP 5
- Thread starter MBH
- Start date
one thing that keeps being repeated, voip telephony is ready for market. 90 percent of the wan networks are not ready for ip trunks.. scrimp on bandwidth, and you might as well give each user a cell phone, spend half the day asking are you still there.. that is the most used phrase on cell phone in america.. the ad's all ask can you here me now, the real world is can you hear me yet?
john poole
bellsouth business
columbia,sc
john poole
bellsouth business
columbia,sc
phoneguy55
Programmer
I think IP trunks are the greatest thing since sliced bread. I just migrated my current employer from 4 digit dialing via the PSTN to IP trunks and the savings are massive. With 10 sights the savings are racking up faster than I can count. I don't know why anyone would do 4 digit dialing via the PSTN. And paying for P2P voice T1's is no longer necessary when you have IP trunking avaiable. The main key to a sucessful deployment is carefull planning and working with good Networking guys. QoS is also a must!!!
If you want up time, go with a hybrid system that does not use the network from the system to the desk phone. Do not use an unreliable operating system for your voice processing(WINDOWS). Use the network for the desk top applications and site to site calling. If you need off network soft phones for remote users, then do it, but QoS is always going to be key issue, put the money in your network to support it or do not rely on it. If the vendor does not assess your network, get some loose leaf tobacco, roll it in their proposal and light one end. Inhale deeply, and make sure you have copies of the proposal to roll more smokes when it goes down and you are stressed because your job depends on the voice performance.
If you go with an open source supplier remember that support is only as good as the supplier, open source suppliers need to support everyones product, and this can spread them thin. $400 per handset is not so bad, if it works, and is supported when it has issues. $100 per handset is not a bargain if they do not work, or they do not support their product, or integrating it with every product it will be used with.
If you go with an open source supplier remember that support is only as good as the supplier, open source suppliers need to support everyones product, and this can spread them thin. $400 per handset is not so bad, if it works, and is supported when it has issues. $100 per handset is not a bargain if they do not work, or they do not support their product, or integrating it with every product it will be used with.
aarenot, I have to say your statements are confusing. Either the whole network is good enough to support voice, or it isn't. What good is a hybrid if the site links are flaky? If the WAN/POTS interface is good enough, then chances are the LAN is as well, just given the fact that most LANs have orders of magnitude more bandwidth, and the fact that they're directly under the control of the company.
And I think we all get it -- you hate Windows, and thus Callmanager.
And I think we all get it -- you hate Windows, and thus Callmanager.
I do not hate windows, I use it frequebtly, and it is adequate for the machine I use to access this forum. It is not even close to adequate to process my voice communications for mission critical uninterupted service. I do not use it for such.
I do not either hate call manager, but it will become adequate when Cisco migrates away from a windows baesd OS for voice processing. Cisco shares this opinion about the windows based voice processing, as they are changing away from the windows OS in the future. In their experience they have come to the same conclusion about this OS. So, I guess that Cisco hates the OS, because it has caused them to change the call manager to another OS due to its performance issues. The cost to Cisco would justify them hating the windows OS, personaly, I do not hate it as I do not use it for functions beryond its capabilities, or reliabilities. I like it, but I am not trying to use a stapler to drive a nail, which I equate to using windows OS for voice processing. It can do it, but it is not designed or equipped to do so on an ongoing, reliable, efficient, stable manner. I like my stapler, it staples. I like my hammer, it drives nails. I like my windows OS, it processes data. I like my voice services, they do not use windows OS.
Call manager is hindered by windows, that is an issue of windows, not call manager. When they get off windows OS, they will become an exponentialy better product for voice processing.
That will allow call manager to be judged in the same class as non-windows based voice processing. Until then, it is not an apples to apples comparison, due to performance, security, and stability issues with the OS.
"What good is a hybrid if the site links are flaky?" It can place calls via the telco to other sites. This is also true if the LAN is down or has issues. I understand the LAN is under the control of the company, do you consider the LAN as reliable as the hybrid off LAN infrastructure? If no, then we have some agreement, if yes, then I disagree with your assesment.
I do not either hate call manager, but it will become adequate when Cisco migrates away from a windows baesd OS for voice processing. Cisco shares this opinion about the windows based voice processing, as they are changing away from the windows OS in the future. In their experience they have come to the same conclusion about this OS. So, I guess that Cisco hates the OS, because it has caused them to change the call manager to another OS due to its performance issues. The cost to Cisco would justify them hating the windows OS, personaly, I do not hate it as I do not use it for functions beryond its capabilities, or reliabilities. I like it, but I am not trying to use a stapler to drive a nail, which I equate to using windows OS for voice processing. It can do it, but it is not designed or equipped to do so on an ongoing, reliable, efficient, stable manner. I like my stapler, it staples. I like my hammer, it drives nails. I like my windows OS, it processes data. I like my voice services, they do not use windows OS.
Call manager is hindered by windows, that is an issue of windows, not call manager. When they get off windows OS, they will become an exponentialy better product for voice processing.
That will allow call manager to be judged in the same class as non-windows based voice processing. Until then, it is not an apples to apples comparison, due to performance, security, and stability issues with the OS.
"What good is a hybrid if the site links are flaky?" It can place calls via the telco to other sites. This is also true if the LAN is down or has issues. I understand the LAN is under the control of the company, do you consider the LAN as reliable as the hybrid off LAN infrastructure? If no, then we have some agreement, if yes, then I disagree with your assesment.
What good is a hybrid if the site links are flaky?" It can place calls via the telco to other sites.....
This is a problem that many people have, not a line down, but a POOR line, just like isdn.
If your isdn is cracking, the PBX will still place calls over it. If your IP trunks are poor, it will still place calls over it. What is required are, good, well mangaed lines, ISDN or otherwise.
I personally would never have a system that is pure IP , just like we're moving from pure ISDN. You choose the best for that site, and at least have some form of back up.
That's my opinion anyway.
Stu..
Only the truly stupid believe they know everything.
Stu.. 2004
This is a problem that many people have, not a line down, but a POOR line, just like isdn.
If your isdn is cracking, the PBX will still place calls over it. If your IP trunks are poor, it will still place calls over it. What is required are, good, well mangaed lines, ISDN or otherwise.
I personally would never have a system that is pure IP , just like we're moving from pure ISDN. You choose the best for that site, and at least have some form of back up.
That's my opinion anyway.
Stu..
Only the truly stupid believe they know everything.
Stu.. 2004
phoneguy55
Programmer
Here is why a hybrid set up is great and this is why I use it where I work. I use the WAN to transport calls to/from each PBX. This is makes for free calls. From the PBX the call is sent to the digital phone via copper cat 3 cable.
IMHO the major cost savings is the transport between PBX A and PBX B. Once the call has reached it's destination, why does it still need to be IP and "clog up" the lan. It doesn't. What are you saving??? The cost of 1 cat 3 cable to each desk. Which is nothing...
Also, If you have LAN/WAN problems. My route patterns will automatticly fail over to the PSTN. So that the end user doesn't even know there is a problem. If you are all IP and you have a problem on the LAN, your phone becomes a paper weight.
IMHO the major cost savings is the transport between PBX A and PBX B. Once the call has reached it's destination, why does it still need to be IP and "clog up" the lan. It doesn't. What are you saving??? The cost of 1 cat 3 cable to each desk. Which is nothing...
Also, If you have LAN/WAN problems. My route patterns will automatticly fail over to the PSTN. So that the end user doesn't even know there is a problem. If you are all IP and you have a problem on the LAN, your phone becomes a paper weight.
I agree with almost all of your comments, but I totally disagree with your cabling statement:
As for the LAN, I think it depends on what your topology looks like. If users are already choking the LAN, then adding voice may not be a good idea.
There are definatly benefits to a hybrid approach (which is what I am propsing for my migration strategy), with new locations being IP only. But what about the cost of a PBX vs. an IP system? I can show close to a $7,000 cost savings opening a new location on an IP-PBX vs. and tradional TDM switch. That does not even speak to the recurring cost that come with a TDM solution. I know I could still go TDM to the desktop, and IP for gateway traffic (PBX to PBX), but then the cost goes up, not down.
The best thing I have seen out of this thread so far is the numerous opinions and options people present. I only hope everyone will see that the devil is in the details, and to focus only on the details that are important to your network. One man's pain is another man's opprotunity.
Scott M.
In my analysis, I found that a location we just opened would have seen a 25% cost reduction in cabling had I removed the Cat3 and only used Cat5. This includes the locations where only a phone is needed, like a lobby, fileroom, lunchroom, etc.
As for the LAN, I think it depends on what your topology looks like. If users are already choking the LAN, then adding voice may not be a good idea.
There are definatly benefits to a hybrid approach (which is what I am propsing for my migration strategy), with new locations being IP only. But what about the cost of a PBX vs. an IP system? I can show close to a $7,000 cost savings opening a new location on an IP-PBX vs. and tradional TDM switch. That does not even speak to the recurring cost that come with a TDM solution. I know I could still go TDM to the desktop, and IP for gateway traffic (PBX to PBX), but then the cost goes up, not down.
The best thing I have seen out of this thread so far is the numerous opinions and options people present. I only hope everyone will see that the devil is in the details, and to focus only on the details that are important to your network. One man's pain is another man's opprotunity.
Scott M.
Scott,
I see your point about the hybrid structered cabling costs.
For pennies a foot difference, I would invest the money in the IP structured wiring on site, and run the hybrid over that. That too may drive the cost up, but it does have its benefits for the converged networks solutions.
I might go with analog Stations on an IP hybrid system, would have low deskset costs. It also adds the desktop phone managers to replace the costly multibutton display desksets. Feature sets would be available via short codes. If the LAN is down use the analog station, if the WAN is down use the PSTN off the hybrid.
I have installed , and serviced IP-PBX, hybrid, PBX, key, even 1A2 service only, as well as structured cabling, and project management for all, so I see it from my own perspectives. They may not be from other than my own traditional communications seven nines mentality.
I also like the hybrid OS is not windows.
You do not always get what you pay for, but you never get what you do not pay for.
I see your point about the hybrid structered cabling costs.
For pennies a foot difference, I would invest the money in the IP structured wiring on site, and run the hybrid over that. That too may drive the cost up, but it does have its benefits for the converged networks solutions.
I might go with analog Stations on an IP hybrid system, would have low deskset costs. It also adds the desktop phone managers to replace the costly multibutton display desksets. Feature sets would be available via short codes. If the LAN is down use the analog station, if the WAN is down use the PSTN off the hybrid.
I have installed , and serviced IP-PBX, hybrid, PBX, key, even 1A2 service only, as well as structured cabling, and project management for all, so I see it from my own perspectives. They may not be from other than my own traditional communications seven nines mentality.
I also like the hybrid OS is not windows.
You do not always get what you pay for, but you never get what you do not pay for.
What about the multiple devices on the desk? I also agree that Cat5 vs. Cat3 is a small difference, but if I can eliminate the Cat3 all together, the cost savings is much bigger. One run to the desk vs. two is a big savings in materials and in some efforts labor (only half the cables to terminate).
I have also worked in this field for 10+ years, and did a VoIP implementation 5 years ago using Nortel ITG cards for trunk side only. (That project failed due to some vendor factors, thus the 2nd attempt now) Again, I agree that 90% of your savings come from the PBX to PBX cost elimination. I also agree the TDM OS (VxWorks and others) is much more stable than any Server OS.
I suppose the real question here is what are the soft benefits that drive users to accept the less than five 9's of service? Extension Mobility, XML phones, managability, presence, what? I am only asking, because for me, cost is the biggest factor, but not everyone I interact with sees just the bottom line.
I do like your last line, kind of like Gretzky saying "You will always miss 100% of the shots you never take." (Apologies to Mr. Gretzky for most likely butchering his actual quote)
Scott M.
I have also worked in this field for 10+ years, and did a VoIP implementation 5 years ago using Nortel ITG cards for trunk side only. (That project failed due to some vendor factors, thus the 2nd attempt now) Again, I agree that 90% of your savings come from the PBX to PBX cost elimination. I also agree the TDM OS (VxWorks and others) is much more stable than any Server OS.
I suppose the real question here is what are the soft benefits that drive users to accept the less than five 9's of service? Extension Mobility, XML phones, managability, presence, what? I am only asking, because for me, cost is the biggest factor, but not everyone I interact with sees just the bottom line.
I do like your last line, kind of like Gretzky saying "You will always miss 100% of the shots you never take." (Apologies to Mr. Gretzky for most likely butchering his actual quote)
Scott M.
Multiple devices on the desktop may be justifiable if the voice communications work because of it. That is not a big bottom line discussion. The reality of the non-windows based OS solution(TDM) versus a windows based OS solution, is a seven nines solution versus a five nines solution. The hybrid (TDM)is going to offer a reliability which is more than likely seven nines.
In order to calculate the bottom line, one must calculate the costs of down time for voice. Then, calculate the cost of 0.001 % downtime versus 0.00001 % down time. When you have the cost of down time computed, the rest is easy, if you are only looking for the bottom line.
The savings in cable to run one cable to each location is apparent, the labor to pull two cables to a location versus one is minimal. Terminating is again easy to conceive the labor savings.
"I suppose the real question here is what are the soft benefits that drive users to accept the less than five 9's of service?"
My response, stop using the five 9's analogy as it does not support a cost based analysis. The cost is calculated in down time, not up time. What is the cost of greater than 0.001 down time? The non TDM solutions state 0.001% down time as their goal, if they meet that goal. TDM, is more like 0.00001% as a goal. Compare those numbers at your cost of down time, then at a performance failure of twice the acceptable down time, 0.002 non-TDM, and 0.00002 for TDM. 100 times non-TDM 0.1 %, TDM 0.001 % down time.
As far as the hybrid, it offers the same aspects for the IP communications, soft phones, etc, etc, etc, as the IP only solution. It simply allows TDM as an option for TDM structured wiring equipped operations.
Soft benefits? I see none of any magnitude.
You do not always get what you pay for, but you never get what you do not pay for.
In order to calculate the bottom line, one must calculate the costs of down time for voice. Then, calculate the cost of 0.001 % downtime versus 0.00001 % down time. When you have the cost of down time computed, the rest is easy, if you are only looking for the bottom line.
The savings in cable to run one cable to each location is apparent, the labor to pull two cables to a location versus one is minimal. Terminating is again easy to conceive the labor savings.
"I suppose the real question here is what are the soft benefits that drive users to accept the less than five 9's of service?"
My response, stop using the five 9's analogy as it does not support a cost based analysis. The cost is calculated in down time, not up time. What is the cost of greater than 0.001 down time? The non TDM solutions state 0.001% down time as their goal, if they meet that goal. TDM, is more like 0.00001% as a goal. Compare those numbers at your cost of down time, then at a performance failure of twice the acceptable down time, 0.002 non-TDM, and 0.00002 for TDM. 100 times non-TDM 0.1 %, TDM 0.001 % down time.
As far as the hybrid, it offers the same aspects for the IP communications, soft phones, etc, etc, etc, as the IP only solution. It simply allows TDM as an option for TDM structured wiring equipped operations.
Soft benefits? I see none of any magnitude.
You do not always get what you pay for, but you never get what you do not pay for.
I have been involved in Many Vendor implementations. Including Cisco Nortel and Avaya. The most problems are actually with the Nortel solutions although not all. Cisco and Avaya also have their issues as well. I have found if the pre install work is done well very few surprises come up. As for reliability I can say that they all have issues at times. With Cisco the most issues I have found have been either gateway or Voice Mail issues. I have never had a CallManager Cluster go completely down so no Voice service was available. I wish I could say the same for Nortel. Yes Windows OS has issues but again this is also managable and has never been a real issue. Look for the features that are important to you and then match them to the vendor. once you have the best vendor look at the implementer as this may be the reason I hear about so many issues.
I would say, what any phone system vendor is selling is the system engineer, technician, project manager, installer, service staff. The equipment brand is not relevant if these people are not quality people. Quality people are not generaly the lowest paid, and therefore not working for the lowest bidder. If you are buying a Jaguar, do not buy it from the corner used car lot, and do not have a back yard mechanic service it. Find the best trained, most experienced firm you can afford. Get references from satisfied customers, talk to the technical staff before signing the contract. Remember, if you find a vendor that charges less than everyone else, they know what their services are worth, and there is a reason why they can afford to offer a lower price.
You do not always get what you pay for, but you never get what you do not pay for.
You do not always get what you pay for, but you never get what you do not pay for.
- Status
Similar threads
- Question