It's hard to guess definitively, but I'd start there. Ideally, sticking it above the roofline would be even better.
-forrest On Fri, Oct 3, 2014 at 4:56 PM, Bill Prince via Af <[email protected]> wrote: > So you think moving it a foot or so out from the wall might help? > > bp > > On 10/3/2014 3:33 PM, Forrest Christian (List Account) via Af wrote: > > There are several rules in the current GPS chipset about whether or not it > can produce sync. If you have 4 they can definitely produce sync. Less > than 4 is somewhat hit or miss depending on the exact orientation of the > satellites, and often more miss than hit. I don't really have control > over this algorithm so I can't be more specific. > > If you can move that one slightly so it's getting on average 1-2 more > sats tracked, then your problem should go away. > > -forrest > > > On Fri, Oct 3, 2014 at 2:38 PM, Bill Prince via Af <[email protected]> wrote: > >> Monitoring Sync Events works. Had to wait 5 days for it to happen, but >> the counter increment corresponds to a loss of sync we had yesterday. >> >> I found this in the AP event log (event of interest in blue): >> >> 09/27/2014 : 07:24:19 PDT : : Bridge Core : Loss of sync pulse from Power >> Port! No other sync source available. >> 09/27/2014 : 07:24:23 PDT : : Bridge Core : Acquired sync pulse from >> Power Port. >> 10/02/2014 : 12:21:45 PDT : : Bridge Core : Loss of sync pulse from Power >> Port! No other sync source available. >> 10/02/2014 : 12:21:51 PDT : : Bridge Core : Acquired sync pulse from >> Power Port. >> >> We also monitor visible/tracked satellites on that SiteMonitor. >> Interestingly, the satellites tracked at about that time was 4 (see marked >> up graph below). I suppose it's possible that the tracked satellites went >> to zero one minute (or less) after the SNMP poll, but it seems rather weird. >> >> Correct me if I'm wrong, but I thought the SiteMonitor only needed one >> satellite to maintain timing after it acquired a 3D fix? >> >> >> bp >> >> On 9/29/2014 10:12 PM, Forrest Christian (List Account) via Af wrote: >> >> Yes that value will increment when the injector detects a loss of sync >> and also when it's restored. >> >> These are definitely good values to monitor, and I know at least one >> customer which does as you suggest and monitors for a non zero value and >> resets the value to zero to clear the error. >> On Sep 29, 2014 7:23 PM, "Bill Prince via Af" <[email protected]> wrote: >> >>> Yeah. Not sure why I thought the index name was where I would get the >>> value. The OID that shows in the UI for the Satellites Visible is: >>> >>> .1.3.6.1.4.1.32050.2.1.28.2.1 >>> >>> The OID for the actual value is >>> >>> .1.3.6.1.4.1.32050.2.1.28.5.1 >>> >>> >>> So I was able to fix that part. What I'm wondering is how to know that >>> We've had a loss in sync. There is something under Binary I/O called "1PPS >>> Active". >>> >>> Seeing as we only poll once every 5 minutes, catching that going to zero >>> seems slim to none. However, I am intrigued by the "Events" value. Does >>> that increment every time the Syncpipe loses sync? In which case, I can >>> zero it out, and set a threshold for whenever it is non-zero (see below). >>> >>> >>> I may try that. >>> >>> >>> bp >>> >>> On 9/29/2014 1:28 PM, Forrest Christian (List Account) via Af wrote: >>> >>> A little out of order: >>> >>> On the OID's .. you may have the wrong OID. There is an oid for the >>> title strings, and an oid for the value. You may want to check the oid you >>> are using. In addition, on the strings tab, there *are* strings which >>> list the specific statellite and signal strength of all of the sats it is >>> receiving a signal from. >>> >>> One more troubleshooting item is the 'pulse received' counter on the >>> analog tab. It should increment once and exactly once per second. I've >>> had good luck comparing this value over a specific time. I.E. at exactly >>> 10 minutes, there should be exactly 600 more pulses. >>> >>> As far as fixing it: I'd move the syncpipe, then try a different >>> one. If a second does the same thing, then we need to look at what else >>> might be causing it. >>> >>> If you want to send in screenshots to [email protected] of the >>> boolean/analog/string tabs from the sitemonitor, I might be able to see >>> something. >>> >>> -forrest >>> >>> >>> On Sep 29, 2014 1:40 PM, "Bill Prince via Af" <[email protected]> wrote: >>> >>>> >>>> One of our many locations where we're using a Packetflux sync >>>> pipe/injector seems to be losing satellite lock once every few days. >>>> Typically it loses it for 2 to 4 seconds, but I've seen at least once where >>>> it went 13 seconds. >>>> >>>> I've not been able to get useful information from the SiteMonitor >>>> because the satellites tracked/Visible OIDs are returning a string with >>>> "Sats in View" and "Num Sats Used" instead of the actual values. (is that a >>>> bug or what? This is on F/W "Jul 29 2012"). >>>> >>>> However, I'm getting messages like this in the AP logs: >>>> >>>> 09/21/2014 : 07:49:00 PDT : : Bridge Core : Loss of sync pulse from >>>> Power Port! No other sync source available. >>>> 09/21/2014 : 07:49:04 PDT : : Bridge Core : Acquired sync pulse from >>>> Power Port. >>>> 09/23/2014 : 18:49:37 PDT : : Bridge Core : Loss of sync pulse from >>>> Power Port! No other sync source available. >>>> 09/23/2014 : 18:49:41 PDT : : Bridge Core : Acquired sync pulse from >>>> Power Port. >>>> 09/23/2014 : 18:49:55 PDT : : Bridge Core : Loss of sync pulse from >>>> Power Port! No other sync source available. >>>> 09/23/2014 : 18:49:59 PDT : : Bridge Core : Acquired sync pulse from >>>> Power Port. >>>> 09/24/2014 : 18:47:15 PDT : : Bridge Core : Loss of sync pulse from >>>> Power Port! No other sync source available. >>>> 09/24/2014 : 18:47:28 PDT : : Bridge Core : Acquired sync pulse from >>>> Power Port. >>>> 09/27/2014 : 07:24:18 PDT : : Bridge Core : Loss of sync pulse from >>>> Power Port! No other sync source available. >>>> 09/27/2014 : 07:24:20 PDT : : Bridge Core : Acquired sync pulse from >>>> Power Port. >>>> >>>> Not sure what I might do here. This is with all the equipment up >>>> against a concrete wall, so there is only a 180 degree view of the sky. >>>> Maybe a little bit less than that because the wall is not flat, maybe about >>>> 170 degree view of the sky. >>>> >>>> The APs are PMP450, and rarely get a GPS lock on the internal GPS. >>>> Maybe I can try moving the sync pipe away from the wall or something. >>>> >>>> >>>> -- >>>> bp >>>> >>>> >>> >> > >
