Temporary buffer overflow for thread [sdrtrunk polyphase buffer processor] throwing away samples

Stephen Leander

unread,

May 24, 2022, 3:34:51 PM5/24/22

to sdrtrunk

I have been running the alpha6 version for over a year, I have the system scheduled to reboot every 24 hours because it will stop working much beyond that. Lately, it appears to work, but there is no audio, so I thought I'd try the newer versions.

With both Beta2 and Beta3, I get "Temporary buffer overflow for thread [sdrtrunk polyphase buffer processor] throwing away samples" errors, and the audio is garbled. Looks like the buffer is 1G?

System is an HP Z420 with 16G memory, and an Airspy receiver running Ubuntu 20.04 LTS

Any suggestions?

Thanks!

wemana...@gmail.com

unread,

May 26, 2022, 10:06:42 PM5/26/22

to sdrtrunk

On Tuesday, May 24, 2022 at 3:34:51 PM UTC-4 slea...@gmail.com wrote:

I have been running the alpha6 version for over a year, I have the system scheduled to reboot every 24 hours because it will stop working much beyond that

No Linux system should need to be rebooted every 24 hours, matter of fact in the Linux community its a BIG BIG DEAL to have up time, IN YEARS... and post and brag about it... so I will do so pfftt! :) ;) 373 days for one box... some others I have had to do some RAM upgrades, SSD installs, so they have low up times.

What stops SDRT or you have an issue with Linux?????

If the issue is SDRT then you can kill and restart the SDRT process not the whole box... via BASH script, crontab, and auto start of channels in SDRT.

Linux issues... have you?

disabled baloo file indexing
disabled unattended upgrades - this updates the apt cache in the background and IS A RESOURCE HOG! Spiking CPU usage!

. Lately, it appears to work, but there is no audio, so I thought I'd try the newer versions.

With both Beta2 and Beta3, I get "Temporary buffer overflow for thread [sdrtrunk polyphase buffer processor] throwing away samples" errors, and the audio is garbled. Looks lik

e the buffer is 1G?

System is an HP Z420 with 16G memory, and an Airspy receiver running Ubuntu 20.04 LTS

Any suggestions?

How active is the system??? Phase I or Phase II ???

JMBE 1.09 or better on the B3 ????

Ok. the HP is a 8 core box... that should handle a pretty decent setup... or well, can be up to 8 cores WHICH PROCESSOR?:

You can find via:

$ sudo dmidecode -t processor

OUTPUT:
# dmidecode 3.2
Getting SMBIOS data from sysfs.
SMBIOS 2.7 present.

Handle 0x0004, DMI type 4, 42 bytes
Processor Information
Socket Designation: CPUSocket
Type: Central Processor
Family: FX
Manufacturer: AMD
ID:
Signature: Family 21, Model 2, Stepping 0
Flags:
Version: AMD FX(tm)-8300 Eight-Core Processor
Voltage: 1.2 V
External Clock: 200 MHz
Max Speed: 3300 MHz
Current Speed: 3300 MHz
Status: Populated, Enabled
Upgrade: Socket AM3
L1 Cache Handle: 0x0005
L2 Cache Handle: 0x0006
L3 Cache Handle: 0x0007
Serial Number: To Be Filled By O.E.M.
Asset Tag: To Be Filled By O.E.M.
Part Number: To Be Filled By O.E.M.
Core Count: 8
Core Enabled: 8
Thread Count: 8
Characteristics:
64-bit capable

Intel® Xeon® Processor E5-2687W 3.1 20 MB 1600 MHz 8 Y Y 3, 7

Intel® Xeon® Processor E5-2665 2.4 20 MB 1600 MHz 8 Y Y 4, 7

Intel® Xeon® Processor E5-1660 3.3 15 MB 1600 MHz 6 Y Y 3, 6

Intel® Xeon® Processor E5-1650 3.2 12 MB 1600 MHz 6 Y Y 3, 6

Intel® Xeon® Processor E5-1620 3.6 10 MB 1600 MHz 4 Y Y 2, 3

Intel® Xeon® Processor E5-1607 3.0 10 MB 1066 MHz 4 N Y N/A

Intel® Xeon® Processor E5-1603 2.8 10 MB 1066 MHz 4 N Y N/A

WHICH USB ports??? BLUE 3.0 or REGULAR 2.0 ones?

What setting for Bandwidth and WHICH AIRSPY???? The Mini does 3 or 6MHz and the RSP (?) does 2 and 10MHz respectively.

Did you use EARLIER VERSIONS WITH OUT ISSUES?

Did you perform the Vector SIMD calibrations?

Is this BOX DEDICATED TO SDRT??? Or are you trying to do other daily tasks or any other tasks on it?????

Lets FIRST try one quick hack...

IF You are using the AirSpy in a BLUE CONNECTOR 3.0 port SWAP To 2.0? Same results?

If same, lets move on to a couple other hacks....

in the bin directory of your SDRT tree, there is a BASH file sdr-trunk we need to EDIT IT as follows:

ORIGINAL
DEFAULT_JVM_OPTS='""--add-exports=javafx.base/com.sun.javafx.event=ALL-UNNAMED" "--add-modules=jdk.incubator.vector" "--add-exports=java.desktop/com.sun.java.swing.plaf.windows=ALL-UNNAMED"'

ADD the "-Xmx5120M" INSIDE THE FIRST SINGLE QUOTE

EDITED:
DEFAULT_JVM_OPTS='"-Xmx5120M" "--add-exports=javafx.base/com.sun.javafx.event=ALL-UNNAMED" "--add-modules=jdk.incubator.vector" "--add-exports=java.desktop/com.sun.java.swing.plaf.windows=ALL-UNNAMED"'

This gives the JVM 5GB ram, for your case, honestly I would actually probably use: 8192M to give it 8GB.....

Restart SDRT, does this polyphase buiffer error reoccur???

IF SO

Goto User Preferences and change the TUNER CHANNELIZER *** FROM **** Polyphase TO >>> HETRODYNE <<<< RESTART SDRT

Does this error reoccur???

IF SO

CHANGE BACK TO POLYPHASE and CHANGE THE Mini from 6Mhz to 3Mhz, and the RSP (?) to 2Mhz RESTART SDRT

Error reoccur?

If the last one fixes, it the box can't keep up with that setting of BANDWIDTH, and I'd bet its the 10MHz on the RSP2 ....

Another thing to look at which MIGHT BE AN ISSUE???

Have you DISABLED baloo file indexing??

Have you DISABLED UNATTENDED UPDATES?? apt via systemd (pfft) will update the APT cache data and its a RESOURCE HOG! As it the baloo file indexer ESPECIALLY if you have HIGH LEVEL LOGGING IN SDRT enabled, ie: decoded traffic and control channel logging...

Report findings and we will go from there......

Stephen Leander

unread,

May 27, 2022, 12:57:21 PM5/27/22

to sdrtrunk

I run lots of other linux machines, and am familiar with not having to reboot. :-) When I set this up, I was unable to figure out why SDRTrunk would stop functioning, appeared to be a memory leak or similar, and just having it restart was the solution. The machine and OS is fine - alpha6 just would not stay up for more than a few days. Given that it was alpha code, I wasn't expecting too much. I wanted to see if the newer code would work, and I ran into the problem in the subject.

The Z420 is 4 core, it's a retired dispatch console from our system. IIRC, I am using the USB2 port as the USB3 port gave me issues (just confirmed with lsusb and usb-devices) It's at a remote site, my simulcast prime site, which is collocated with one of the 6 simulcast RF sites. I run the P25 system for the area, and this supplies official feeds. All 800Mhz Phase I, 18 channels. I only have it decoding 3 Talkgroups, and streaming them to Broadcastify. I have VNC and SSH access to maintain it.

I can run it to failure, but not sure how much troubleshooting the Alpha6 code should get... I'll check on the rest when I get a chance, it's running the alpha6 code right now and works better than the Beta3. I might be able to troubleshoot over the weekend or early next week.

Thanks for the input and a list of things to try!

Steve

From dmseg:

SMBIOS 2.7 present.
[ 0.000000] DMI: Hewlett-Packard HP Z420 Workstation/1589, BIOS J61 v03.96 10/29/2019
[ 0.000000] tsc: Fast TSC calibration using PIT
[ 0.000000] tsc: Detected 2793.071 MHz processor

From dmidecode

# dmidecode 3.2
Getting SMBIOS data from sysfs.
SMBIOS 2.7 present.

Handle 0x0004, DMI type 4, 42 bytes
Processor Information

Socket Designation: CPU0
Type: Central Processor
Family: Xeon
Manufacturer: Intel
ID: D7 06 02 00 FF FB EB BF
Signature: Type 0, Family 6, Model 45, Stepping 7
Flags:
FPU (Floating-point unit on-chip)
VME (Virtual mode extension)
DE (Debugging extension)
PSE (Page size extension)
TSC (Time stamp counter)
MSR (Model specific registers)
PAE (Physical address extension)
MCE (Machine check exception)
CX8 (CMPXCHG8 instruction supported)
APIC (On-chip APIC hardware supported)
SEP (Fast system call)
MTRR (Memory type range registers)
PGE (Page global enable)
MCA (Machine check architecture)
CMOV (Conditional move instruction supported)
PAT (Page attribute table)
PSE-36 (36-bit page size extension)
CLFSH (CLFLUSH instruction supported)
DS (Debug store)
ACPI (ACPI supported)
MMX (MMX technology supported)
FXSR (FXSAVE and FXSTOR instructions supported)
SSE (Streaming SIMD extensions)
SSE2 (Streaming SIMD extensions 2)
SS (Self-snoop)
HTT (Multi-threading)
TM (Thermal monitor supported)
PBE (Pending break enabled)
Version: Intel(R) Xeon(R) CPU E5-1603 @ 2.80GHz
Voltage: 1.1 V
External Clock: 100 MHz
Max Speed: 3800 MHz
Current Speed: 2800 MHz
Status: Populated, Enabled
Upgrade: Socket LGA2011

L1 Cache Handle: 0x0005
L2 Cache Handle: 0x0006
L3 Cache Handle: 0x0007

Serial Number: Not Specified
Asset Tag: Not Specified
Part Number: Not Specified
Core Count: 4
Core Enabled: 4
Thread Count: 4
Characteristics:
64-bit capable
Multi-Core
Execute Protection
Enhanced Virtualization
Power/Performance Control

wemana...@gmail.com

unread,

May 27, 2022, 11:34:37 PM5/27/22

to sdrtrunk

On Friday, May 27, 2022 at 12:57:21 PM UTC-4 slea...@gmail.com wrote:

I run lots of other linux machines, and am familiar with not having to reboot. :-) When I set this up, I was unable to figure out why SDRTrunk would stop functioning, appeared to be a memory leak or similar, and just having it restart was the solution. The machine and OS is fine - alpha6 just would not stay up for more than a few days. Given that it was alpha code, I wasn't expecting too much. I wanted to see if the newer code would work, and I ran into the problem in the subject.

rather than reboot the machine, I'd just run... this in a BASH script under crontab ... rebooting a box just because of a program unless its a hardlock.. freaks me out...

kill -9 $(pidof java | awk '{print $1}')

sleep 10

/path/to/sdrtrunk/bin/sdr-trunk

You may even want to setup things to give time for another cron job to run as root
sudo sysctl -w vm.drop_caches=2
sudo sysctl -w vm.drop_caches=1
sudo sysctl -w vm.drop_caches=3

To clear out memory.... this takes a little bit of working out to kill SDRT as the user, wait for root to run this job, and then the user script restart SDRT.. Can be done, probably take about 3-4 minites.. since you don't want to run SDRT as root and the sysctl needs root to run...

The Z420 is 4 core, it's a retired dispatch console from our system. IIRC, I am using the USB2 port as the USB3 port gave me issues (just confirmed with lsusb and usb-devices) It's at a remote site, my simulcast prime site, which is collocated with one of the 6 simulcast RF sites. I run the P25 system for the area, and this supplies official feeds. All 800Mhz Phase I, 18 channels. I only have it decoding 3 Talkgroups, and streaming them [ed:censored]. I have VNC and SSH access to maintain it.

I have issues using USB3 ports with some SDR's, namely the box that does my main production in similar circumstances...

I think this is going to boil down to this:

#1 Memory
#2 PROCESSOR v. SDR BANDWIDTH.

Based on my hunches from doing this on SDRT since about 0.3??? Or so???

My main production box runs 0.5.0.**ALPHA 6*** namely because it needs to be up 24/7/365. And I've not had the maintenance time period to do the swap out needed, YET... Although based on some things in re BETA 3.. I may hold off...

Now.. this box is a 4 CORE AMD A10 based unit, it runs on my master zone the busiest of the 7 zones (each county is a zone) this runs using a 2 SDR's from a HYDRA SDR 4 SDR units.. I'd like to use the other 2 for V/UHF NBFM channels if possible under BETA 3.. but I am thinking that may not fly based on some posts here...

This zone is 24 channels. Phase I....and I feed... way way way more than 3 talk groups, WAY MORE. There are 5 feeds for LE alone, with about 4 TG's each... The FD/EMS feed is only the dispatch side, 2 TG's as I find FD/EMS a snooze zone.... The PRIVATE FG feeds is 15 TG's and if the calls hump, you will be hard pressed to keep up...I listen I find a call of interest in the page outs etc... The HEMS feed is 8 TG's... as it feeds the HEMS ops channels, LZ OPs TG, and then some other stuff is in there for me like some of the CW HAIL inter agency stuff... Another PRIVATE feed is the HEAR for this zone.. It basically takes the 8 HEAR TG"s for the hospitals in the zone and feeds it. Light but steady use...and I can get the details on the "good" calls... when I want them. I also have SOME things recorded basically my personal LE zone is recorded for my own PERSONAL PRIVATE INSTANT RECALL RECORDER. The HEMS groups are recorded again, for my own private review when I can't listen live....

The two important things here is 16GB RAM (max for this MB) and I give SDRT 8GB RAM for the HEAP... AND I only need 2 SDR's for the 24 channels, fortunately we didn't get boned in Rebanding, another system is a bane of my existence they got royally hosed in Rebanding and got channels from 861 into 852! I let it run on a PI3 via OP25, since they have the Secret Stasi LE, so it only can feed FD over 10 TG's... OP25 fits it well.... I will combine all of this in the near future, so I can pick off some other feeds on there as well they are low volume, but have a place...

What I am getting at do you ABSOLUTELY NEED to run the Airpsy at a large 6 or 10Mhz setting to get all your frequencies???? If not can you run them at the 3 setting and possibly supplement with another SDR at 2.4Mhz to get the outliers??? I do that on another system which is not related to mine, AirSpy at 3Mhz,and a V3 at 2.4..Once this one comes online fully, that won't fly due to the frequency spacing ..its another thorn in my side for this an an another issue....it runs on a twin of the main box AMD A10 based unit... 16GB ram, 8GB to the heap.. For the 1 FEED and 10 TG's of that feed, its not a big deal, right now.. in the future it will be 4-6 feeds spread out over about 40ish TG's depending on how things fly... this one though is Phase II.

I can run it to failure, but not sure how much troubleshooting the Alpha6 code should get... I'll check on the rest when I get a chance, it's running the alpha6 code right now and works better than the Beta3. I might be able to troubleshoot over the weekend or early next week.

I don''t think the issue is CODE in ALPHA 6 in your situation. The tuner error you are getting is 99.99999999999% HARDWARE of the box running SDRT... As in inadequate to run the SDR at 6 or 10MHz bandwidth...

Basically this going to come down to:

Are you willing and able to put 8GB More RAM into that box, be it cost and/or authorized to upgrade

If you absolutely need the 6/10Mhz settings of the SDR. Then you are going to need the 8 Core CPU for this box that upgrade is $60-150 depending on where you want to get the chip from... The ONLY WAY I can run Airpsys at 6/10 or my HackRF ones at greater than 4-5MHz is via my Ryzen 5 or my 8 Core AMD boxen.... and can you upgrade ie: authorized and/or cost(s)...

FAIR WARNING that using the USB bus on some systems can NOT KEEP UP with Airpspys and HRFO's at high bandwidths, and USING A DEDICATED USB CARD which has a better USB controller and goes via the PCI bus, which the MB does too, but the chips in the MB are not high end USB controllers, is about $80-100 depending on where you get it... it does give you 4 DEDICATED CHANNEL USB PORTS.

If you review my main setup, what I am doing on a very under powered box, AMD A10 is down to : 2 2.4MHz SDR"s via the HydraSDR...That 4.8Mhz is more than enough for my 24 channels...and all that recording and feeding...

Just to compare:
https://www.cpubenchmark.net/compare/AMD-A10-4655M-APU-vs-Intel-Xeon-E5-1603/11vs1209

How does this work??? 4.8MHz SDR bandwidth.... the same applies to the twin using an airspy, but because the airspy's filters are pretty disastrous to the edge channels its only about 70% of the 3Mhz.. or about 2.1Mhz in reality.. theres a thread on the GitHub issues which listrs what the finally settings are for these... its pretty abysmal, but this filtering comes in handy in the 700MH block v.s the D Block FirstNet pollution.

You MIGHT get away with the 8GB RAM upgrade AND CHANGING TO HETRODYNE PROCESSOR.. which I've found does tend to work a little better on lower powered boxes at 3/6Mhz settings... but the only way to know is to test it...

Honestly I've found that 8GB is pretty much the "sweet spot" for HEAP size with SDRT regardless of what every else you do. From my extensive testing and tweaking on things I've got a pretty good knowledgeabase built up on this, and its pretty much going to come down to what I've posted.. the tests basically will just bear out where you might be able to tweak things to get by or not...

charley....@gmail.com

unread,

May 28, 2022, 8:46:57 AM5/28/22

to sdrtrunk

What does HTOP say?

Stephen Leander

unread,

May 31, 2022, 6:57:52 PM5/31/22

to sdrtrunk

HTOP indicates that CPU wise, it's pretty busy, but under 4 with the 4 cores, Load average (15min) 3.44. Memory wise, I only have 2.9G currently out of 15.6G available.

My simulcast RX frequency spread is 7.1375Mhz. I can dedicate another receiver if that will help, but I have no problem purchasing a better machine for this task either - I took a machine slated for disposal (auction) to see if this would work. This was a proof of concept - it will allow me to dynamically add talkgroups with no incremental cost - if I can get it working properly. :-) I can change the CPU as well, I did it to a few other of the Z420's, but I have budget to buy something a bit more capable as well.

Where can I control or allocate more memory?

wemana...@gmail.com

unread,

Jun 1, 2022, 1:45:19 AM6/1/22

to sdrtrunk

On Tuesday, May 31, 2022 at 6:57:52 PM UTC-4 slea...@gmail.com wrote:

HTOP indicates that CPU wise, it's pretty busy, but under 4 with the 4 cores, Load average (15min) 3.44. Memory wise, I only have 2.9G currently out of 15.6G available.

16GB RAM - VGAFB RAM = 15.9, and you have 2.9G FREE RAM out of this 15.6GB??? What else is running????????

That doesn't seem to add up in my experience with RAM usage for Linux...

My simulcast RX frequency spread is 7.1375Mhz.

7.1375 / 2.4MHz Standard SDR BW = 2.97, so 3 STANDARD SDR's... now since you are using the Airspy RSP it has 90% of the 10MHz BW, setting so 9MHz, so you are good, with some breathing room, on the Mini, not going to work 100%, you will get TUNER UNAVAILABLE during peak loading, 6MHz x 0.83 = 4.98MHz... (see https://github.com/DSheirer/sdrtrunk/issues/815 as to why these are used )

So you need 10MHz from the Airpsy... which is going to lead head first into the Tuner Buffer issue on USB saturation issues...

I can dedicate another receiver if that will help, but I have no problem purchasing a better machine for this task either - I took a machine slated for disposal (auction) to see if this would work. This was a proof of concept - it will allow me to dynamically add talkgroups with no incremental cost - if I can get it working properly. :-) I can change the CPU as well, I did it to a few other of the Z420's, but I have budget to buy something a bit more capable as well.

Here in lies the issue:

How much $$$ to test v. gain in resolving this... ie: CPU upgrade, RAM upgrade, SDR's, USB card for dedicated use...

Do you by chance have 3 V3 or similar xDVB-T SDR dongles that can be used for testing.. the idea here is to use 3 SDR V3 or similar at 2.4MHz and run it on the box.. does that work?

I would try with 2, and then add the 3rd, as you likely will run into the USB saturation point with 3, possibly.

How much "budget" do you have to test, and play... to figure out the resolution... BUT... here is what I am going to guess on this...

Getting high BW SDR's to work takes
a) CPU heft
b) USB support
c) RAM

None of the HackRF or Airpsys at 6,10+MHz work at anything less than my Ryzen 5 3500U or the 8 Core box... thats using just the standard USB ports on them...

You can try this incrementally,

So personally for you I would get a MIN of 32GB of RAM IN there, and DEDICATE 8GB to SDRT as per below...

A DEDICATED USB CONTROLLER CARD:
https://www.amazon.com/StarTech-com-Express-SuperSpeed-Dedicated-Channels/dp/B00HJZEA2S

And RETEST with the Airspy...

The ONLY WAY to FIX the USB SATURATION ISSUE is:

1) CPU HEFT
2) USB Controllers dedicated for EACH CHANNEL, no shared ports on the BUS like MB ones...

The RAM helps make things work for SDRT and JVM.... BUT its PURE CPU PROCESSING AND USB separation that solves this tuner issue 99.9999% of the time...

This alone may solve things... possibly... if not then its a CPU upgrade...

CPU UPGRADE to a MIN 8 CORE UNIT

Where can I control or allocate more memory?

in your SDRT tree, in the bin directory, there is an sdr-trunk BASH file, edit it:

You need to use the JAVA OPTION Xmx to allocate memory to the Java HEAP

-Xmx8192M sets it to 8GB

DEFAULT_JVM_OPTS='""--add-exports=javafx.base/com.sun.javafx.event=ALL-UNNAMED" "--add-modules=jdk.incubator.vector" "--add-exports=java.desktop/com.sun.java.swing.plaf.windows=ALL-UNNAMED"'

CHANGE TO:
DEFAULT_JVM_OPTS='"-Xmx8192M" "--add-exports=javafx.base/com.sun.javafx.event=ALL-UNNAMED" "--add-modules=jdk.incubator.vector" "--add-exports=java.desktop/com.sun.java.swing.plaf.windows=ALL-UNNAMED"'

Stephen Leander

unread,

Jun 1, 2022, 12:11:05 PM6/1/22

to sdrtrunk

2.9G in use - not free, sorry if that wasn't clear. The system has plenty of free memory. I have 2 Airspy's at work, I do have some of the cheap dongles at home, I used them for various things. One of them has been doing POCSAG decoding for almost 10 years now on an original Raspberry Pi.

wemana...@gmail.com

unread,

Jun 2, 2022, 9:01:12 AM6/2/22

to sdrtrunk

On Wednesday, June 1, 2022 at 12:11:05 PM UTC-4 slea...@gmail.com wrote:

2.9G in use - not free, sorry if that wasn't clear. The system has plenty of free memory. I have 2 Airspy's at work

Ok.. not how I read it, clearly... use the -Xmx8192 option as I listed in the other post... RAM for the JVM HEAP Is a big cure to many things. the BETA 2-3 seem to use less of the available heap RAM, but A6 definitely will suck it all in at once, and for it, thats a good thing 8GB or more.....

, I do have some of the cheap dongles at home, I used them for various things.

Grab a few of them if they are not in use... take them in to that box and see if you get the error... start with 2, and you likely will get TUNER NOT AVAILABLE for some calls on just 2 during peak demand... if you have data calls on there too, may as well disable them, and not waste the time with them... that may make the tuner unavailable go down, to near 0, since you can't do much with most of it any way.....

This is sort of an incremental thing...

Test with 2 SDR V3 types, OK??? No errors? ADD 3rd... ERRORS or USB LOCK UP (RED LINE IN TUNERS TAB)??? BINGO!

If you get that 3 V3 types work, if you add a 4th and try to decode another channel or something that probably comes close to triggering it... but either way, if 3 V3 like SDR's work the issue becomes

USB saturation....

The resolution to that is:

1) CPU - Possibly better CPU can clear the USB bus faster, thus avoiding this.... $90-127 depending on chip source.

2) USB ports on board --- replace USB Ports with DEDICATED CONTROLLER USB ports $99 - The board I listed, and has been proven to work, theres a long thread on this where theres some one decoding 10+ channels or something...

I get away with using my AMD 4 Core SoC setup as I only need 2 SDR's to do the 24 channels, and I dump data grants which I can't do anything with any way... so... adding in the other 2 for NBFM, probably won't work on this setup... so I am not even going to bother with it for now...I've went years with that split for some V/UHF stuff, so no big deal.

If 3 V3 like SDR's work thats a pretty good indication for CPU upgrades... as the 4 core can't keep up with the data....The USB card change would be the last change I make...as CPU and computational speed is the bigger issue in this error... especially if are NOT GETTING THE RED LINE IN THE TUNER TAB (USB ERROR)... you didn't mention it so, I will take the overflow is the only error... so thats CPU heft, and the RAM for the heap... Java's heap and garbage collection process is known, well, garbage. :)

One of them has been doing POCSAG decoding for almost 10 years now on an original Raspberry Pi.

I'd do my POCSAG verifiier on a Pi if I free one up, right now its on something else, mainly as the LOG is HUGE.... for similar length... there was designs for more, but some things got in the way.. which is another reason its not on a Pi....ie: doing various logging functions.. it becomes a need to have a Pi with more than just an SD card for memory, hence why its on a regular box with a HD....

Reply all

Reply to author

Forward