Hardware: Polycom HDX 7000 HD Serial: 880952105784CN
Software: Release - 18.104.22.168_1-3960 Date: 2009-11-02 09:55:02-0600 Built By: buildmaster Basis Stream: HDX2506_integration
Linux Name: BusyBox v1.1.2
Linux version: 22.214.171.124.p2.9
gcc version 4.1.0) #1 PREEMPT
Wed Mar 25 16:18:36 CDT 2009
Camera: Eagle Eye MPTZ-6
Part #: 1624-23412-00...(illegible)
Camera system firmware: 
Camera firmware: 
Camera lens firmware: 
Camera FPGA firmware: 
The camera video stream on this old hardware has started malfunctioning:
1. The system boots, dialing screen loads, with Eagle Eye video live in background.
2. After about 30 seconds, the live background video freezes
3. After a few minutes, the system suddenly reboots
After poking around in the diagnostic logs, it looks like the codec brainbox doesn't notice when the camera video stream is lost. It remains blissfully unaware of any problems... I can pan the camera around with the video frozen, or manipulate the connect menu, and the video remains frozen.
But when the codec brainbox decides to go into sleep mode (default 3 minutes of inactivity), it tells the camera to spin around to the sleep position, and then tries to turn off the video stream, ..... but discovers it can't find the video stream .... then a watchdog timer says "proc 0 stopped responding" and forces a reset.
Is there anything that I can do about this without sending it in for repair?
I already tried the soft configuration reset (hold reset 15 seconds at main menu), and the hard operating system reload (hold reset while turning on), and neither of these help.
Is it the Eagle Eye camera that is the problem, and not the codec brainbox? I checked the cable from the camera to the codec, and there are no bent or broken pins.
Due to the weird design of the IR receiver being in the camera, but not on the codec front, it is hard to use the codec brainbox without the camera connected.
Also it is unclear to me if the codec brainbox can "tolerate" running by itself standalone without an Eagle Eye camera plugged into it.
Or, perhaps the problem is the SanDisk 512meg compact flash drive? Is it developing rotten bits after being in use for 8 years? I remember 10+ years ago reading about how early flash media was not exactly stable, and after many years the electric charge can dissipate from the storage cells.
I see that a "full factory reset" actually doesn't reset everything on the flash drive. The diagnostic logs "Support information package" still contains all the previous error and message logs from before the "full OS reset". So if there are any corrupt files with rotted bits lurking in the flash, they were not all rewritten fresh, and corrupt data may still be present.
Then also, this system has really old software, 126.96.36.199_1-3960 .... I suppose this video freezing could be a bug that is fixed by a newer OS version. When I look up the serial number, it says we are entitled for v2.6 and v3.0, but I can't find those listed in the downloads for the HDX 7000, only 3.1 is available.
welcome back to the Polycom Community.
A couple of your old posts => here <= are still open / pending as you have not marked these as "Accept as a solution" or at least provided some form of feedback or answer.
If they are in this state nobody finding them via a community search will know if an answer or advice provided was useful and has maybe helped you.
Could you therefore kindly go over them and mark or answer as appropriate ?
If they are marked as "Accept as a solution" other users can find these easier and it helps them to utilize the community more efficiently.
Some to these ancient software versions are harder to find but please check => here <=
You should be able to find a 3.0.x software as you already stated that you cannot upgrade to 3.1.x
Please ensure to provide some feedback if this reply has helped you so other users can profit from your experience.
Polycom Global Services
This project is on hold while I wait for the delivery of some brand new flash memory cards that are identical to what is installed in the HDX 7000, in case the existing flash has gone bad.
SanDisk 512meg Compact Flash - SDCFJ-512
US$25 per 0.5 gig CF card is laughably expensive for such a tiny amount of storage in modern terms.
But the technology is nearly a decade old now, and the codec would probably choke on a cheaper modern SanDisk Ultra 16 gig compact flash which is only US$18 by comparison....
SanDisk Ultra 50 MB/s 16gig Compact Flash - SDCFHS-016G-AFFP
Or, woo! A 32gb CF for only US$25.
SanDisk Ultra 50 MB/s 32gig Compact Flash - SDCFHS-032G-G46
But it's not worth bothering to try this. The codec was designed for 512 meg CF and it probably can't see much beyond that. Compact Flash was originally just a small form factor IDE/(P)ATA hard drive interface, with the same sector addressing limits as physical hard drives of the time.
Well, that was a bit of an adventure, but ultimately it didn't solve anything.
I bought three SanDisk 512meg CF for testing, but apparently only one of them works with the HDX 7000, because for some reason the flash drives are not all exactly the same size. They vary by a few sectors more or less.
The average person probably would not notice this using the CF cards in a camera or multimedia device, but I have found that the HDX 7000 flat out refuses to accept anything but a CF with the exact right number of sectors.
I do not know if this is a normal situation for compact flash media.
But I do know that storage media defects are a fact of life, and it is common for a manufacturer to have to create an error map during production to deal with this for other types of storage media
I took the original CF media from the broken HDX7000, booted CloneZilla, dropped to the command line, located the flash media with fdisk, and then made a perfect sector-level copy of the HDX partition structures.
sudo fdisk -l | less
Disk /dev/sdc: 492.8 MiB, 516612096 bytes, 1009008 sectors
sudo dd if=/dev/sdc of=HDX7000-freezing-backup.dd bs=512 status=progress
Then I wrote this perfect copy, back to the new flash media. It didn't fit at all on the CF media that was a bit too small, and the HDX 7000 refused to boot with the slightly too big media, even though I zeroed the media first to blank out the small amount of slack space beyond the last partition....
sudo dd if=/dev/zero of=/dev/sdc bs=512 status=progress
(wait for "out of disk space" error, which is normal when zeroing a drive this way)
sudo dd if=HDX7000-freezing-backup.dd of=/dev/sdc bs=512 status=progress
But the HDX 7000 did accept and properly boot from the one CF out of the three that fdisk reported with the exact same number of sectors. It still does the video freeze, but that is fine.
I was not really expecting this media duplication to fix anything.... If the problem was CF media corruption, then all this does is copy bad data to new media. But this way at least I can try doing upgrades while holding onto the original CF media in case some other route surfaces to fix this.
We have another HDX 7000 that is still working properly, so after verifying I can successfully boot the dd-duplicated media of the broken hardware, I then moved on to the real purpose... duplicate the media from the working hardware, and then try booting that in the broken hardware.
I will not try the actual good source media from the working HDX 7000 in the broken unit, because I have seen situations where broken hardware corrupts other good hardware that is being used to try to diagnose and fix it.
Result? ... well it seemed promising at first, but nope, the media copy from the working system also froze the video after a while on the malfunctioning unit.
I recognize this may actually be some other problem, due to some sort of configuration mismatch of using the firmware from one unit in the other. But the two HDX were bought at the same time and the MAC addresses are only about 50 units apart so they are very closely matched hardware.
So, looks like this is probably some sort of hardware failure, beyond the flash file system. I have a likely suspect... The mid-to-late 2000s were the era of the "bad capacitor crisis", so it really would not surprise me at all that there are bad caps lurking in either the power supply or the camera.
Probably will pop the shield off the power supply, and look inside the camera to see if there are any bulged or leaked capacitors.
I am going to try upgrading to the 3.x version OS with the duplicated flash media, but I have a strong suspicion that this won't fix the video freezing either.
And I need to see if the "bad" HDX 70000 works properly without the Eagle Eye camera connected.
Also I need to see if there is a way to get into the Linux shell of the flaky Polycom. If the thing is dying, well I suppose I could try fooling around with the shell scripts and diagnostic tools to see if they can tell me anything.
There is not much point spending US$2500 to US$5500 to officially repair this end of life device. That money would be better spent on a new videoconferencing system/brand, or a used but working HDX 7000 from eBay.
I am closing this post as opening your HDX invalidates your warranty and we do not support "homemade" repairs.
Opening a device exposes parts of the power supply which are dangerous for untrained users so therefore we do not support discussions about such within the official Polycom community.
Polycom Global Services