Adding Live Buffer to HD-FOX by UPD

Black Hole

May contain traces of nut
I thought I would have a play with fitting a large UPD to HD1 (and the others of it works out) to enable live pause, and see what the longevity of the UPD is like in this situation. I bought one of these:

eBay: UNBRANDED 32GB USB MEMORY STICK

The first thing was to format it. I plugged it into HDR1 and it was reported as 30.5GB, with format options unavailable. I then plugged it into HD3 (1.02.29/2.15) and used the Telnet menu to format it Ext2. The menus detected it as 31000MiB!

The next thing is to replicate the custom environment I have on the existing 2GB UPD for HD1, rather than have to rebuild it, and that's where I am stuck and could use some advice. The same will also be useful for replicating the environment across multiple HD-FOXes.

I tried plugging the new UPD into HDR1 and using the HD1 WebIF media browser to copy-and-paste /[ModSettings] and /mod across to the (SMB mounted) HDR1 drive2, but although /[ModSettings] seemed to copy (on inspection the long filenames do not report correctly, but that might be a consequence of the mount), /mod would not go at all.

I tried the Telnet command line to do the same thing, but it wouldn't let me. HD1 is not mounted on HDR1, so I couldn't try the other way around.

The next idea I had was to dig out my Linux netbook and plugged in both UPDs. When I tried to copy across with the file manager it said I didn't have enough permissions.

It seems to me it would be very useful to have a Telnet command that can package up the environment into a tar, which could then be transported to another system and untarred.

Help!
 
Install dropbear-ssh and rsync on both then you can sync the environment across with something like:

Code:
humax# rsync -avr /media/drive1/mod/ <ip of other humax>:/media/drive1/mod/

They will, of course, end up with the same host name until you change it back on the target.
 
Code:
humax# rsync -avr /media/drive1/mod/ <ip of other humax>:/media/drive1/mod/
Can this be made to work with a target drive plugged into an HDR, with the rsync command creating the directory (or even better to sync from the root folder)?
 
Well, the rsync command will create the target directory (providing the last argument is finished with a trailing / - it's always best to add trailing slashes when syncing directory to directory with rsync.

You could sync to the drive plugged into the HDR if you wanted - the only problem would be if you had HDR-only packages installed, they would end up on the HD.

You could plug both into the HDR and sync from one to the other or sync from an HD on the network to a drive plugged into the HDR.
 
What level of network support is required at each end? Or is it sufficient that both devices exist on the network, accessed by IP address?
 
What does this mean?:
Code:
HDRFOX3# cd /media/drive2
HDRFOX3# ls
ls: ./mod: Input/output error
ls: ./Video: Input/output error
ls: ./Audio: Input/output error
ls: ./[ModSettings]: Input/output error
epgsavedata  lost+found
 
Code:
HDFOX1# rsync -avr /media/drive1/ 192.168.1.75:/media/drive2/

Host '192.168.1.75' is not in the trusted hosts file.
(fingerprint md5 5e:cd:82:9b:42:ea:0a:9f:fe:40:29:88:e9:6d:a5:68)
Do you want to continue connecting? (y/n) y

ssh: Connection to root@192.168.1.75:22 exited: Didn't validate host key
rsync: connection unexpectedly closed (0 bytes received so far) [sender]
rsync error: error in rsync protocol data stream (code 12) at io.c(601) [sender=3.0.8]
HDFOX1#
 
Well, the rsync command will create the target directory (providing the last argument is finished with a trailing / - it's always best to add trailing slashes when syncing directory to directory with rsync.
Are you sure? I always thought that it would create a target directory if the first argument does not end in a '/'. If the first argument ends in a '/' it will copy the contents of that directory to the destination. AFAIK any trailing '/' is ignored on the last argument.
 
What does this mean?:
Code:
HDRFOX3# cd /media/drive2
HDRFOX3# ls
ls: ./mod: Input/output error
ls: ./Video: Input/output error
ls: ./Audio: Input/output error
ls: ./[ModSettings]: Input/output error
epgsavedata  lost+found
Corrupt file system and/or faulty disk.
 
Tsk. It was created by the Telnet maintenance menu. I'll try again (I'm getting in a knot with this - it would be quicker to make a note of the current setup and recreate it on the new UPD).

Instead of rsync, isn't there a way I can do what I first suggested and create an image file? How about something which taps into the package format?
 
I formatted the "32GB" UPD as exFAT (to test the new exFAT support), and (as formatted) Windows reports:

Used: 196,608 bytes
Free: 32,500,383,744 bytes
Total: 32,500,580,352 bytes

So it looks like it's a marketing 32GB rather than an actual 32GiB (again).
 
I have hit a problem updating HD1 from 1.02.29+2.12 to 1.02.29+2.15 and incorporating the 32GB UPD. Typical, I should know not to change more than one thing at a time, but we all try to take short cuts.

Although I didn't think it was necessary I followed the Wiki instructions and did the prepare_for_upgrade diagnostic followed by loading vanilla 1.02.29, then loaded 2.15. I connected by Telnet and entered Maintenance Mode to convert the UPD to Ext2, and found I had to reformat the UPD from exFAT to FAT32 first. Then I ran the conversion process. OK so far.

Then I rebooted to normal mode and expected to connect to the web loader, but got no response to the IP address. The HD-FOX boots showing "F2.15", the Internet Setting menu shows the right IP address, and Telnet connects to the menu, but no web page.
Code:
humax# service
/bin/sh: service: not found
I would be grateful for a clue what to do next. I tried reloading the CF, but the programming cycle went through very quickly (I guess it does a compare) and made no difference (twice).

I then tried putting the old UPD back and rebooting, and the WebIF reappears - but I want to reload the CF onto the new UPD..
 
I switched sticks over again while the HD-FOX was live, and the "USB device loading" message came up but not followed by the option icons. This time the web request came up Error 404, so I guess that means mongoose is still running from before the switch. Curiously the Telnet server still knows the host name.
Code:
HDFOX1# cd media
HDFOX1# ls drive1
ls: drive1/Video: Input/output error
Audio        epgsavedata  lost+found   mod
HDFOX1# ls drive1/mod
ls: drive1/mod/var: Input/output error
bin   boot  etc   sbin  tmp
HDFOX1# ls drive1/mod/bin
HDFOX1#

Bear in mind this is from a stick that has had nothing done to it other than a format to Ext2 via the Telnet menu and otherwise just been sitting there.

Do the I/O errors suggest there is something wrong with the stick (it formatted exFAT OK and moved an 8GB file) or that the Ext2 conversion didn't work properly? What stops the web downloader page appearing in the absence of the second stage install?

Unfortunately I have to suspend operations for a few days due to other pressing demands. Back to the old stick...
 
When you get back to it, post the output of rag.log. That should give us a clue about why the web service isn't starting.
 
Umm... presumably that means a version of rag.log from when the 32GB UPD is plugged in and I can't get anything sensible?
 
Back
Top