Re: [time-nuts] OT: NTP server questions
If you are rolling your own I would advise a Soekris net4501 (US $173 new) over any netbook for several reasons: cheaper, more rugged, better solution overall. It all comes down to requirements, budget, and who the user is (which drives the first two). If I were doing a system for myself, it's hard to beat what I have with a net4501 ($175) , a Thunderbolt ($120) , a TAPR Clock Block ($62), and a pulse stretcher (TAPR FatPPS $44). I have a 12V power supply (yes it's a switcher, but I can live with that) to drive the +12, +5, and -12 required by the Thunderbolt. Whole thing is powered by 12V, and is very portable. Also extremely accurate. All for about $400 US. Some assembly required. If you don't need that level of accuracy you can get very reasonable solution with a net4501 and a Garmin GPS18xLVC. Power the GPS from a 5V pin off of the Soekris. Total cost less than $250. Under other situations a commercial box is more appropriate. If I were asked to specify a system that I would not be responsible to administer I would recommend a commercial box such as the Symmetricom. Ralph On Nov 30, 2010, at 1:39 AM, Chris Albertson wrote: What do you need to do? What precision is required and how many clients will you be serving. For most normal uses you don't need a special purpose server system. A $600 notebook PC and any GPS unit with a serial connection and a copy of Linux or BSD. On Mon, Nov 29, 2010 at 11:36 AM, Robert Darlington rdarling...@gmail.com wrote: Hi guys, I'm looking to buy an NTP server for a field deployable server system. I currently have a Symmetricom SyncServer S250 which does more than I need. I am considering buying an S200 (same as the S250 but without the ability to connect to an external frequency standard). My gut feeling is I don't need a rubidium oscillator or even an OCXO internally since we'd be going from a power on to being used state in under an hour, outside of any kind of temperature controlled environment. Are there any manufacturers out there besides Symmetricom that I should be looking at for something like this? The unit we're considering buying is about $4k new and is inside of our budget. Thanks, Bob ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. -- = Chris Albertson Redondo Beach, California ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] OT: NTP server questions
Basically I need something to provide time within one second. I can't roll my own in this case. At home I have a Soekris box with a custom built gps board for my normal level of time-nuttery but this is not for home. I'm looking for commercially made rack mount servers that will not have Internet access for reference and will need to rely on gps. The system will need to serve time to less than 100 systems but it will live in a nasty environment in the back of a humvee (or something like an S-250 electronics enclosure) with no air conditioning of any sort. I'm currently looking at Symmetricom, Trimble NetRS, EndRun, and Meinberg. Also, I apologize if I'm breaching protocol with this thread. This is just about the first post I ever saw on this list that didn't get a steady stream of replies. I was actually questioning if this even made it into the list till I saw this reply. Yes this is for work and I could really use some opinions on this stuff since dropping $4-6k isn't in my nature to do at home for a network clock so I never gave this thought before. Thanks again, Bob On Mon, Nov 29, 2010 at 11:39 PM, Chris Albertson albertson.ch...@gmail.com wrote: What do you need to do? What precision is required and how many clients will you be serving. For most normal uses you don't need a special purpose server system. A $600 notebook PC and any GPS unit with a serial connection and a copy of Linux or BSD. On Mon, Nov 29, 2010 at 11:36 AM, Robert Darlington rdarling...@gmail.com wrote: Hi guys, I'm looking to buy an NTP server for a field deployable server system. I currently have a Symmetricom SyncServer S250 which does more than I need. I am considering buying an S200 (same as the S250 but without the ability to connect to an external frequency standard). My gut feeling is I don't need a rubidium oscillator or even an OCXO internally since we'd be going from a power on to being used state in under an hour, outside of any kind of temperature controlled environment. Are there any manufacturers out there besides Symmetricom that I should be looking at for something like this? The unit we're considering buying is about $4k new and is inside of our budget. Thanks, Bob ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. -- = Chris Albertson Redondo Beach, California ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] OT: NTP server questions
At 08:46 AM 11/30/2010, Robert Darlington wrote: Also, I apologize if I'm breaching protocol with this thread. This is just about the first post I ever saw on this list that didn't get a steady stream of replies. I was actually questioning if this even made it into the list till I saw this reply. Yes this is for work and I could really use some opinions on this stuff since dropping $4-6k isn't in my nature to do at home for a network clock so I never gave this thought before. Seems pretty on-topic to me, especially if you get one and test it against your Soekris based rig. -- newell N5TNL ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] OT: NTP server questions
Robert Darlington wrote: Basically I need something to provide time within one second. I can't roll my own in this case. At home I have a Soekris box with a custom built gps board for my normal level of time-nuttery but this is not for home. I'm looking for commercially made rack mount servers that will not have Internet access for reference and will need to rely on gps. The system will need to serve time to less than 100 systems but it will live in a nasty environment in the back of a humvee (or something like an S-250 electronics enclosure) with no air conditioning of any sort. I'm currently looking at Symmetricom, Trimble NetRS, EndRun, and Meinberg. Also, I apologize if I'm breaching protocol with this thread. This is just about the first post I ever saw on this list that didn't get a steady stream of replies. I was actually questioning if this even made it into the list till I saw this reply. Yes this is for work and I could really use some opinions on this stuff since dropping $4-6k isn't in my nature to do at home for a network clock so I never gave this thought before. Thanks again, Bob I've used True-Time (now Symmetricom) boxes for about 10-11 years at work to do time and frequency in lab sorts of environments. You've identified all the big players in the field for off the shelf boxes, really, all you have to do now is just find the one that works for you in terms of options, etc. If this application is a sit in the corner and be a time source then an all-in-one box designed for the need is pretty attractive. Sure, you could scrounge up an old (or new) PC, load software, connect some sort of USB or RS232 interface GPS, etc. But, I suspect that if you're doing it at work, the cost of your time to scrounge, assemble, document, and test would be more than just writing a check for a kilobuck or so to any of the commercial vendors, which would get you a shiny new debugged box with a user manual and a warranty. Of course, if your work is somewhere with free labor (Hey, Bob, for your senior project, how about making a GPS time server) that can change the strategy As does the possibility of used gear.. Not quite as nice as when there were pallet loads of Z3801s for a few hundred bucks each...but stuff does turn up (e.g. all those Thunderbolts) ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] OT: NTP server questions
FWIW, you can't go wrong with Meinberg. Several versions to choose from, reliable, and priced well. Rob Kimberley -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Robert Darlington Sent: 30 November 2010 2:47 PM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] OT: NTP server questions Basically I need something to provide time within one second. I can't roll my own in this case. At home I have a Soekris box with a custom built gps board for my normal level of time-nuttery but this is not for home. I'm looking for commercially made rack mount servers that will not have Internet access for reference and will need to rely on gps. The system will need to serve time to less than 100 systems but it will live in a nasty environment in the back of a humvee (or something like an S-250 electronics enclosure) with no air conditioning of any sort. I'm currently looking at Symmetricom, Trimble NetRS, EndRun, and Meinberg. Also, I apologize if I'm breaching protocol with this thread. This is just about the first post I ever saw on this list that didn't get a steady stream of replies. I was actually questioning if this even made it into the list till I saw this reply. Yes this is for work and I could really use some opinions on this stuff since dropping $4-6k isn't in my nature to do at home for a network clock so I never gave this thought before. Thanks again, Bob On Mon, Nov 29, 2010 at 11:39 PM, Chris Albertson albertson.ch...@gmail.com wrote: What do you need to do? What precision is required and how many clients will you be serving. For most normal uses you don't need a special purpose server system. A $600 notebook PC and any GPS unit with a serial connection and a copy of Linux or BSD. On Mon, Nov 29, 2010 at 11:36 AM, Robert Darlington rdarling...@gmail.com wrote: Hi guys, I'm looking to buy an NTP server for a field deployable server system. I currently have a Symmetricom SyncServer S250 which does more than I need. I am considering buying an S200 (same as the S250 but without the ability to connect to an external frequency standard). My gut feeling is I don't need a rubidium oscillator or even an OCXO internally since we'd be going from a power on to being used state in under an hour, outside of any kind of temperature controlled environment. Are there any manufacturers out there besides Symmetricom that I should be looking at for something like this? The unit we're considering buying is about $4k new and is inside of our budget. Thanks, Bob ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. -- = Chris Albertson Redondo Beach, California ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] OT: NTP server questions
On Tue, Nov 30, 2010 at 6:46 AM, Robert Darlington rdarling...@gmail.com wrote: Basically I need something to provide time within one second. I can't roll my own in this case. I assume you don't have internet access. If you had access only a few times per day you can get better than 1 second using remote servers. It's not the cost but the fact that you will have one more computer to haul around and you add another single point failure to the system. For 1 second accuracy you could simply run NPT servers on one or more of the othercomputers. Run an NTP server process on each of the other servers and plug in a standard USB GPS unit. If you are driving around likely you even have a GPS unit already. Yes it needs to operate in that environment but what can be more robust than zero additional hardware? If the requirement is for no beter than 100 ms there is no need for a dedicated server hardware for NTP. In fact you should run an NTP server process on all your server computers so as not to have a single point of failure.Also you be surprised how well a group of NTP peers maintain time even when cut off from a tier one server. The clients can figure out which of several servers have the best clocks and can do about 1 second per day even with no GPS link So in this case I'd suggest not buying a dedicated NTP server at all. Then yopu have near cost, zero space, zero power and zero chance of failure. -- = Chris Albertson Redondo Beach, California ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] OT: NTP server questions
Rob, you wound me :-) Actually, the Meinberg unit is a great box. The commercial devices pretty much all fit the need that you describe. One caution I would make about them is the note about the environmentals. From experience I can tell you that a HV is a very harsh environment for shock and vib that could, and probably does, sometimes exceed the spec on these boxes. For a box that has to survive repeated power cycles, dusty conditions with limited airflow and unmanned operation, the commercial boxes are optimal but can still have issues. For OP, Rb in these devices is primarily targeted at holdover, providing extended periods of NTP service when the input time source (e.g. GPS) is lost. It doesn't sound like this is a requirement for your application. -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Rob Kimberley Sent: Tuesday, November 30, 2010 7:55 AM To: 'Discussion of precise time and frequency measurement' Subject: Re: [time-nuts] OT: NTP server questions FWIW, you can't go wrong with Meinberg. Several versions to choose from, reliable, and priced well. Rob Kimberley -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Robert Darlington Sent: 30 November 2010 2:47 PM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] OT: NTP server questions Basically I need something to provide time within one second. I can't roll my own in this case. At home I have a Soekris box with a custom built gps board for my normal level of time-nuttery but this is not for home. I'm looking for commercially made rack mount servers that will not have Internet access for reference and will need to rely on gps. The system will need to serve time to less than 100 systems but it will live in a nasty environment in the back of a humvee (or something like an S-250 electronics enclosure) with no air conditioning of any sort. I'm currently looking at Symmetricom, Trimble NetRS, EndRun, and Meinberg. Also, I apologize if I'm breaching protocol with this thread. This is just about the first post I ever saw on this list that didn't get a steady stream of replies. I was actually questioning if this even made it into the list till I saw this reply. Yes this is for work and I could really use some opinions on this stuff since dropping $4-6k isn't in my nature to do at home for a network clock so I never gave this thought before. Thanks again, Bob On Mon, Nov 29, 2010 at 11:39 PM, Chris Albertson albertson.ch...@gmail.com wrote: What do you need to do? What precision is required and how many clients will you be serving. For most normal uses you don't need a special purpose server system. A $600 notebook PC and any GPS unit with a serial connection and a copy of Linux or BSD. On Mon, Nov 29, 2010 at 11:36 AM, Robert Darlington rdarling...@gmail.com wrote: Hi guys, I'm looking to buy an NTP server for a field deployable server system. I currently have a Symmetricom SyncServer S250 which does more than I need. I am considering buying an S200 (same as the S250 but without the ability to connect to an external frequency standard). My gut feeling is I don't need a rubidium oscillator or even an OCXO internally since we'd be going from a power on to being used state in under an hour, outside of any kind of temperature controlled environment. Are there any manufacturers out there besides Symmetricom that I should be looking at for something like this? The unit we're considering buying is about $4k new and is inside of our budget. Thanks, Bob ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. -- = Chris Albertson Redondo Beach, California ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] OT: NTP server questions
Hi Greg (and everyone!) The environment will be dusty and hot and cold and hot and cold... There will be air conditioning but probably not much in the way of filtering. Diesel and gas generators run out of fuel, UPS's are probably not going to be going along with our stuff since they're too heavy. I'm actually pretty concerned about the heat issue getting out of control and we were kidding at work about spontaneous fusion reactions due to the power density. Our #1 problem with this project has been time synchronization. Most of the big issues were worked out but sometimes still crop up. The really big issue we see over and over is when people timestamp data and send that data into our server, the timestamp doesn't meet our spec. We require time in UNIX time in whole numbers of seconds and assume UTC. You'd be surprised at how many months have passed since we first realized people were sending us time in miliseconds so all dates looked like they were 40,000 years in the future. Our computers are good but we don't have anywhere near that kind of predictive capability. The next big issue was when a sensor or some other device sends the data to an intermediary that sends the data to us. More times than not there is a timezone correction done twice. The sensor will send in UTC and somebody will subtract 7 hours to put it into UTC again. They're still doing this from time to time. I recommended we modify our message protocol to include a timezone offset to force people to think about it but that opens a whole other can of worms. Do you do it in seconds or minutes or hours and introduce a whole other potential where people send in the wrong format? Guys, I think I have everything I need now -at least when it comes to mobile time serving. I really appreciate the responses. Thank you. -Bob On Tue, Nov 30, 2010 at 5:35 PM, Greg Dowd gd...@symmetricom.com wrote: Rob, you wound me :-) Actually, the Meinberg unit is a great box. The commercial devices pretty much all fit the need that you describe. One caution I would make about them is the note about the environmentals. From experience I can tell you that a HV is a very harsh environment for shock and vib that could, and probably does, sometimes exceed the spec on these boxes. For a box that has to survive repeated power cycles, dusty conditions with limited airflow and unmanned operation, the commercial boxes are optimal but can still have issues. For OP, Rb in these devices is primarily targeted at holdover, providing extended periods of NTP service when the input time source (e.g. GPS) is lost. It doesn't sound like this is a requirement for your application. -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Rob Kimberley Sent: Tuesday, November 30, 2010 7:55 AM To: 'Discussion of precise time and frequency measurement' Subject: Re: [time-nuts] OT: NTP server questions FWIW, you can't go wrong with Meinberg. Several versions to choose from, reliable, and priced well. Rob Kimberley -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Robert Darlington Sent: 30 November 2010 2:47 PM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] OT: NTP server questions Basically I need something to provide time within one second. I can't roll my own in this case. At home I have a Soekris box with a custom built gps board for my normal level of time-nuttery but this is not for home. I'm looking for commercially made rack mount servers that will not have Internet access for reference and will need to rely on gps. The system will need to serve time to less than 100 systems but it will live in a nasty environment in the back of a humvee (or something like an S-250 electronics enclosure) with no air conditioning of any sort. I'm currently looking at Symmetricom, Trimble NetRS, EndRun, and Meinberg. Also, I apologize if I'm breaching protocol with this thread. This is just about the first post I ever saw on this list that didn't get a steady stream of replies. I was actually questioning if this even made it into the list till I saw this reply. Yes this is for work and I could really use some opinions on this stuff since dropping $4-6k isn't in my nature to do at home for a network clock so I never gave this thought before. Thanks again, Bob On Mon, Nov 29, 2010 at 11:39 PM, Chris Albertson albertson.ch...@gmail.com wrote: What do you need to do? What precision is required and how many clients will you be serving. For most normal uses you don't need a special purpose server system. A $600 notebook PC and any GPS unit with a serial connection and a copy of Linux or BSD. On Mon, Nov 29, 2010 at 11:36 AM, Robert Darlington rdarling...@gmail.com wrote: Hi guys, I'm looking to buy an NTP server for a field
Re: [time-nuts] OT: NTP server questions
Er, rather, add 7 hours. You get the idea. On Tue, Nov 30, 2010 at 6:10 PM, Robert Darlington rdarling...@gmail.comwrote: Hi Greg (and everyone!) The environment will be dusty and hot and cold and hot and cold... There will be air conditioning but probably not much in the way of filtering. Diesel and gas generators run out of fuel, UPS's are probably not going to be going along with our stuff since they're too heavy. I'm actually pretty concerned about the heat issue getting out of control and we were kidding at work about spontaneous fusion reactions due to the power density. Our #1 problem with this project has been time synchronization. Most of the big issues were worked out but sometimes still crop up. The really big issue we see over and over is when people timestamp data and send that data into our server, the timestamp doesn't meet our spec. We require time in UNIX time in whole numbers of seconds and assume UTC. You'd be surprised at how many months have passed since we first realized people were sending us time in miliseconds so all dates looked like they were 40,000 years in the future. Our computers are good but we don't have anywhere near that kind of predictive capability. The next big issue was when a sensor or some other device sends the data to an intermediary that sends the data to us. More times than not there is a timezone correction done twice. The sensor will send in UTC and somebody will subtract 7 hours to put it into UTC again. They're still doing this from time to time. I recommended we modify our message protocol to include a timezone offset to force people to think about it but that opens a whole other can of worms. Do you do it in seconds or minutes or hours and introduce a whole other potential where people send in the wrong format? Guys, I think I have everything I need now -at least when it comes to mobile time serving. I really appreciate the responses. Thank you. -Bob On Tue, Nov 30, 2010 at 5:35 PM, Greg Dowd gd...@symmetricom.com wrote: Rob, you wound me :-) Actually, the Meinberg unit is a great box. The commercial devices pretty much all fit the need that you describe. One caution I would make about them is the note about the environmentals. From experience I can tell you that a HV is a very harsh environment for shock and vib that could, and probably does, sometimes exceed the spec on these boxes. For a box that has to survive repeated power cycles, dusty conditions with limited airflow and unmanned operation, the commercial boxes are optimal but can still have issues. For OP, Rb in these devices is primarily targeted at holdover, providing extended periods of NTP service when the input time source (e.g. GPS) is lost. It doesn't sound like this is a requirement for your application. -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Rob Kimberley Sent: Tuesday, November 30, 2010 7:55 AM To: 'Discussion of precise time and frequency measurement' Subject: Re: [time-nuts] OT: NTP server questions FWIW, you can't go wrong with Meinberg. Several versions to choose from, reliable, and priced well. Rob Kimberley -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Robert Darlington Sent: 30 November 2010 2:47 PM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] OT: NTP server questions Basically I need something to provide time within one second. I can't roll my own in this case. At home I have a Soekris box with a custom built gps board for my normal level of time-nuttery but this is not for home. I'm looking for commercially made rack mount servers that will not have Internet access for reference and will need to rely on gps. The system will need to serve time to less than 100 systems but it will live in a nasty environment in the back of a humvee (or something like an S-250 electronics enclosure) with no air conditioning of any sort. I'm currently looking at Symmetricom, Trimble NetRS, EndRun, and Meinberg. Also, I apologize if I'm breaching protocol with this thread. This is just about the first post I ever saw on this list that didn't get a steady stream of replies. I was actually questioning if this even made it into the list till I saw this reply. Yes this is for work and I could really use some opinions on this stuff since dropping $4-6k isn't in my nature to do at home for a network clock so I never gave this thought before. Thanks again, Bob On Mon, Nov 29, 2010 at 11:39 PM, Chris Albertson albertson.ch...@gmail.com wrote: What do you need to do? What precision is required and how many clients will you be serving. For most normal uses you don't need a special purpose server system. A $600 notebook PC and any GPS unit with a serial connection and a copy of Linux or BSD
Re: [time-nuts] OT: NTP server questions
But, I suspect that if you're doing it at work, the cost of your time to scrounge, assemble, document, and test would be more than just writing a check for a kilobuck or so to any of the commercial vendors, which would get you a shiny new debugged box with a user manual and a warranty. Of course, if your work is somewhere with free labor (Hey, Bob, for your senior project, how about maki Most sysadmins that I know can add NTP to some handy server in a few minutes. That's for getting the time over the net. Yes, it would take longer to add a GPS system. It's probably more work to supervise a senior project than it is to do it yourself. -- These are my opinions, not necessarily my employer's. I hate spam. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] OT: NTP server questions
Basically I need something to provide time within one second. I can't roll my own in this case. At home I have a Soekris box with a custom built gps board for my normal level of time-nuttery but this is not for home. I'm looking for commercially made rack mount servers that will not have Internet access for reference and will need to rely on gps. The system will need to serve time to less than 100 systems but it will live in a nasty environment in the back of a humvee (or something like an S-250 electronics enclosure) with no air conditioning of any sort. I'm currently looking at Symmetricom, Trimble NetRS, EndRun, and Meinberg. 100 systems is falling off a log. If you have some other server that isn't overloaded, it may be better overall to run NTP on that system. The idea is to avoid adding another box just for NTP. If you do something like that, you may still need a GPS unit. I don't have any good suggestions. One alternative to a GPS unit would be to set the time somehow, and then just let it coast for the duration of your experiment. You could set it by hand, or via a temporary network connection, or from the RTC/TOY clock at boot time (but then you have to check that occasionally) or ... That's adding an operational step in order to get simpler hardware. -- These are my opinions, not necessarily my employer's. I hate spam. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] OT: NTP server questions
That's exactly what we've done in the past (setting it when on the network and letting the clock do what it wants) and that's fine. The actual time isn't as important as the agreement on what time it is. This is certainly the cheaper way to go and is becoming a viable option. I agree about the 100 systems. Everything we're considering is way overkill but it gives us some simplicity that the end user wants. -Bob On Tue, Nov 30, 2010 at 7:09 PM, Hal Murray hmur...@megapathdsl.net wrote: Basically I need something to provide time within one second. I can't roll my own in this case. At home I have a Soekris box with a custom built gps board for my normal level of time-nuttery but this is not for home. I'm looking for commercially made rack mount servers that will not have Internet access for reference and will need to rely on gps. The system will need to serve time to less than 100 systems but it will live in a nasty environment in the back of a humvee (or something like an S-250 electronics enclosure) with no air conditioning of any sort. I'm currently looking at Symmetricom, Trimble NetRS, EndRun, and Meinberg. 100 systems is falling off a log. If you have some other server that isn't overloaded, it may be better overall to run NTP on that system. The idea is to avoid adding another box just for NTP. If you do something like that, you may still need a GPS unit. I don't have any good suggestions. One alternative to a GPS unit would be to set the time somehow, and then just let it coast for the duration of your experiment. You could set it by hand, or via a temporary network connection, or from the RTC/TOY clock at boot time (but then you have to check that occasionally) or ... That's adding an operational step in order to get simpler hardware. -- These are my opinions, not necessarily my employer's. I hate spam. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] OT: NTP server questions
I had suggested the same thing. In fact I'd argue not having an NTP box is more reliable than having one. A non-esistant box can't fail. But don't run just one NTP server, run one on every non-overloaded server. You clients will automatically sync with whichever server is best On Tue, Nov 30, 2010 at 6:09 PM, Hal Murray hmur...@megapathdsl.net wrote: 100 systems is falling off a log. If you have some other server that isn't overloaded, it may be better overall to run NTP on that system. The idea is to avoid adding another box just for NTP. If you do something like that, you may still need a GPS unit. -- = Chris Albertson Redondo Beach, California ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] OT: NTP server questions
That won't work in my application. I can't run anything on any server but one I provide specifically for time, which is why I'm looking at dedicated time servers. Believe me when I say this crossed my mind and was crossed off the list. Just about every system is MS Windows based which means one server to sync against unless I add complexity (add NTP) and that won't work in my application since I only control a few of the systems and we don't want anything else running but our custom software. I can't dictate to the other participants in the project that they run software and even if I did they'd all end up doing it differently and I'd be back to one guy using UTC, another TAI. Guys, I appreciate the comments and now it's just a matter of deciding what hardware to buy. I like my existing Symmetricom box but don't want to re-purpose it for this project so it means buying a 2nd one. They're all roughly $4k starting so it's coming down to what color I like best ;-) -Bob On Tue, Nov 30, 2010 at 11:32 PM, Chris Albertson albertson.ch...@gmail.com wrote: I had suggested the same thing. In fact I'd argue not having an NTP box is more reliable than having one. A non-esistant box can't fail. But don't run just one NTP server, run one on every non-overloaded server. You clients will automatically sync with whichever server is best On Tue, Nov 30, 2010 at 6:09 PM, Hal Murray hmur...@megapathdsl.net wrote: 100 systems is falling off a log. If you have some other server that isn't overloaded, it may be better overall to run NTP on that system. The idea is to avoid adding another box just for NTP. If you do something like that, you may still need a GPS unit. -- = Chris Albertson Redondo Beach, California ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] OT: NTP server questions
rdarling...@gmail.com said: That's exactly what we've done in the past (setting it when on the network and letting the clock do what it wants) and that's fine. The actual time isn't as important as the agreement on what time it is. This is certainly the cheaper way to go and is becoming a viable option. How long does it take for data to get from way out in the field to your system? Earlier, you said that you only needed time to within a second. That's a long time for networking. If it's only 100 ms (pulled out of the air) for the data to get from the sensor to your system, then forget all the timestamps as the local system and just use the arrival time at your system. If a significant part of the delay is things like RS-232 baud rates, you can correct for that constant offset. (or semi-constant if the length of a message varies, but maybe you can record the length of the message and do the right correction) -- These are my opinions, not necessarily my employer's. I hate spam. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] OT: NTP server questions
There are sometimes delays up to 30 minutes or so due to processing of sensor data till it makes it into my system which is also way out in the field. Imagine a shipping pallet full of equipment that gets air dropped into the middle of nowhere. That IS my network, and it has no connection to anything but the sensors deployed with it. Maybe on a good day I'll have a wireless or satellite connection back to another network that may or may not have a time server on it. I can't ever assume this will be the case. I will, however, be rotating hard disks full of collected sensor data for off-line analysis (forensic stuff) at a less remote location. I will have spare parts (including NTP servers) at this location that will be visited daily. I really only need time to one second. Most file systems don't have any finer granularity and we're not making scientific measurements. We have a system to display transient events inside of a time window that can be arbitrarily set, but typically no wider than 15 minutes. If the data is 15 minutes old, it's not important anymore. If the data is 40,000 years in the future due to a bad time stamp like when it's marked in miliseconds since the epoch (unix time) instead of seconds, we won't ever see it since it too is outside of my window. If it's 4 seconds old, that's okay. I'll see it inside of my time window. The problem with arrival times is that things I'm looking for don't happen the instant I get alerted. I may want to go back and look at when events happened and compare that with when I was notified so I can see where the bottle necks are. If I have a 5 second delay and 5 seconds of clock drift in the same direction, I don't see any bottleneck which is why I need to know roughly what time it is at each sensor, computer, and server at least to one second. It's rare when something I'm looking for comes directly into one of my systems. Almost always there is some processing going on by other teams on the project that are working independently from my team trying to make use of some of the data in different ways. They would then generate the alert message into my system. One of these teams that created data used a TAI reference clock and the folks using this data used UTC (as do I) and it really caused problems since one set of sensors is saying some event happened at a particular time and yet, there was nothing there at that time according to other sensors. I think in this case it was cars driving down a street (I have no idea how they actually determine this since I just ingest data) and another group took video data to generate tracks for vehicles that were not there at the time the vehicle detectors said they were. 34 second delay (TAI vs UTC) means I'm more than a half mile away from the sensor if I'm going about 60mph. If I'm trying to predict when a car will be near another sensor based on bad data, we'll never be looking at the right time. -Bob On Wed, Dec 1, 2010 at 12:08 AM, Hal Murray hmur...@megapathdsl.net wrote: rdarling...@gmail.com said: That's exactly what we've done in the past (setting it when on the network and letting the clock do what it wants) and that's fine. The actual time isn't as important as the agreement on what time it is. This is certainly the cheaper way to go and is becoming a viable option. How long does it take for data to get from way out in the field to your system? Earlier, you said that you only needed time to within a second. That's a long time for networking. If it's only 100 ms (pulled out of the air) for the data to get from the sensor to your system, then forget all the timestamps as the local system and just use the arrival time at your system. If a significant part of the delay is things like RS-232 baud rates, you can correct for that constant offset. (or semi-constant if the length of a message varies, but maybe you can record the length of the message and do the right correction) -- These are my opinions, not necessarily my employer's. I hate spam. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] OT: NTP server questions
I think we are getting precision and accuracy confused, And they are quite different, In the time-nut world we use them as synonyms, In reality its possible to have a precise measurement which is not accurate (think instrument with bad timebase it's repeatable and reads out to a high degree of precision but it's INACCURATE). We can also have a accurate measurement in with resolution in gross units like minutes or seconds but be highly ACCURATE but have a low degree of PRECISION. What's needed is ACCURATE time referenced to a external standard to within milliseconds with a resolution of 1 second as application does not understand any unit smaller than 1 second. Scott Sent from my Verizon Wireless BlackBerry -Original Message- From: Robert Darlington rdarling...@gmail.com Sender: time-nuts-boun...@febo.com Date: Wed, 1 Dec 2010 00:34:00 To: Discussion of precise time and frequency measurementtime-nuts@febo.com Reply-To: Discussion of precise time and frequency measurement time-nuts@febo.com Subject: Re: [time-nuts] OT: NTP server questions There are sometimes delays up to 30 minutes or so due to processing of sensor data till it makes it into my system which is also way out in the field. Imagine a shipping pallet full of equipment that gets air dropped into the middle of nowhere. That IS my network, and it has no connection to anything but the sensors deployed with it. Maybe on a good day I'll have a wireless or satellite connection back to another network that may or may not have a time server on it. I can't ever assume this will be the case. I will, however, be rotating hard disks full of collected sensor data for off-line analysis (forensic stuff) at a less remote location. I will have spare parts (including NTP servers) at this location that will be visited daily. I really only need time to one second. Most file systems don't have any finer granularity and we're not making scientific measurements. We have a system to display transient events inside of a time window that can be arbitrarily set, but typically no wider than 15 minutes. If the data is 15 minutes old, it's not important anymore. If the data is 40,000 years in the future due to a bad time stamp like when it's marked in miliseconds since the epoch (unix time) instead of seconds, we won't ever see it since it too is outside of my window. If it's 4 seconds old, that's okay. I'll see it inside of my time window. The problem with arrival times is that things I'm looking for don't happen the instant I get alerted. I may want to go back and look at when events happened and compare that with when I was notified so I can see where the bottle necks are. If I have a 5 second delay and 5 seconds of clock drift in the same direction, I don't see any bottleneck which is why I need to know roughly what time it is at each sensor, computer, and server at least to one second. It's rare when something I'm looking for comes directly into one of my systems. Almost always there is some processing going on by other teams on the project that are working independently from my team trying to make use of some of the data in different ways. They would then generate the alert message into my system. One of these teams that created data used a TAI reference clock and the folks using this data used UTC (as do I) and it really caused problems since one set of sensors is saying some event happened at a particular time and yet, there was nothing there at that time according to other sensors. I think in this case it was cars driving down a street (I have no idea how they actually determine this since I just ingest data) and another group took video data to generate tracks for vehicles that were not there at the time the vehicle detectors said they were. 34 second delay (TAI vs UTC) means I'm more than a half mile away from the sensor if I'm going about 60mph. If I'm trying to predict when a car will be near another sensor based on bad data, we'll never be looking at the right time. -Bob On Wed, Dec 1, 2010 at 12:08 AM, Hal Murray hmur...@megapathdsl.net wrote: rdarling...@gmail.com said: That's exactly what we've done in the past (setting it when on the network and letting the clock do what it wants) and that's fine. The actual time isn't as important as the agreement on what time it is. This is certainly the cheaper way to go and is becoming a viable option. How long does it take for data to get from way out in the field to your system? Earlier, you said that you only needed time to within a second. That's a long time for networking. If it's only 100 ms (pulled out of the air) for the data to get from the sensor to your system, then forget all the timestamps as the local system and just use the arrival time at your system. If a significant part of the delay is things like RS-232 baud rates, you can correct for that constant offset. (or semi-constant if the length of a message
[time-nuts] OT: NTP server questions
Hi guys, I'm looking to buy an NTP server for a field deployable server system. I currently have a Symmetricom SyncServer S250 which does more than I need. I am considering buying an S200 (same as the S250 but without the ability to connect to an external frequency standard). My gut feeling is I don't need a rubidium oscillator or even an OCXO internally since we'd be going from a power on to being used state in under an hour, outside of any kind of temperature controlled environment. Are there any manufacturers out there besides Symmetricom that I should be looking at for something like this? The unit we're considering buying is about $4k new and is inside of our budget. Thanks, Bob ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] OT: NTP server questions
What do you need to do? What precision is required and how many clients will you be serving. For most normal uses you don't need a special purpose server system. A $600 notebook PC and any GPS unit with a serial connection and a copy of Linux or BSD. On Mon, Nov 29, 2010 at 11:36 AM, Robert Darlington rdarling...@gmail.com wrote: Hi guys, I'm looking to buy an NTP server for a field deployable server system. I currently have a Symmetricom SyncServer S250 which does more than I need. I am considering buying an S200 (same as the S250 but without the ability to connect to an external frequency standard). My gut feeling is I don't need a rubidium oscillator or even an OCXO internally since we'd be going from a power on to being used state in under an hour, outside of any kind of temperature controlled environment. Are there any manufacturers out there besides Symmetricom that I should be looking at for something like this? The unit we're considering buying is about $4k new and is inside of our budget. Thanks, Bob ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. -- = Chris Albertson Redondo Beach, California ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.