DSM-G600, DNS-3xx and NSA-220 Hack Forum

Chumby wrote:

Thanks Dan,

So I have done the funplug and can telnet in. 

Can you give me an example line where I can delete everything off Volume2, and then copy everything from Volume1 to Volume2 (lh drive to rh drive) using rsync.

Much appreciated!

Chumby

One way to delete everything on Volume2 is the following:

cd /mnt/HD_b2
rm -r *

To use rsync, you'll need to get the binary if you haven't already.  If if the location of the binary is in your path, you can do the following (what Skydreamer posted earlier in this thread)

rsync -rlptDv --delete  source_dir destination_dir >log_file

Enter rsync -h, for an explanation of what each command line parameter does.

Chumby wrote:

Thanks for that Dan.  Will try and find rsync and try the command above...

Appreciate the help.

Cheers
Chumby

It's available in the download section of the wiki page.  Fonz's fun_plug also includes it.

If you mean to make a copy from one disk of the DNS to the other one, using the windows scheduler for this is not the best solution, as the DNS can do it by itself, using "cron" (I think Emacs already mentionned this earlier).
- no client needs to be up or connected
- you can schedule the copy at a time (late at night?) when you are more or less sure there is no other activity on the DNS. Although rsync is very fast, it's always better to avoid making backups and modifying files at the same time.

- put your "manual" commands in a shell script
- run the script once to check it works (be carefull with the path, it's a good idea to set the path manually in the script).
- put the script on the crontab.
    You're done...

Use rync, as explained above, first manually, to check it suits your needs... Find the rsync parameters that suit you the best ("man rsync...").

Once you're happy with your "rsync command", put the command line in a shell script.  Put the script in the appropriate place on the disk, give it 755 permissions.
In the script, prefer to use to absolute path to the rsync executable, so you don't have to bother with the path when run by cron.
Once the script works fines, use cron to run it at the period you want (have a look at "man cron" or "man crontab", I don' remember. Maybe both).
Usually, the way to add something to cron is to use "crontab -e", but I did not test it on the DNS. This is standard Unix/Linux. To run the script in cron, again, give the absolute path to the script. This should work.

I just posted a result from my experience with RAID1.

My system basically stopped recognizing the array thus making the data inaccessible.
Then Linux wont read it because of the RAID1 configuration, so your data is locked up on the drives.
Then I had to use a utility to recover.

I am now simply manually syncing the data as standard drives and using a USB drive as an off line backup.

So I would not use the RAID1, I have over 230GB of photos what a disappointment that would have been.

Bob

Forum link:
http://dns323.kood.org/forum/t926-RAID1 … isted.html

Last edited by det1rac (2007-09-15 01:41:00)

Any attempt to format disk #2 with disk #1 removed (or vice versa) will result in the RAID settings being changed and the unit will not rebuild the array.

Here's how I would do it

1) Switch off DNS-323 and remove the disk of your choice
2) Switch on the DNS-323 and verify the availability of the test data
3) Switch off the DNS-323
4) Install the "replacement" disk - this needs to be the same size or larger than the original, and "clean", ie no partitions or data.
5) Switch on the DSN-323 and follow the prompts to format and resync
6) After the resync is complete, verify the availability of data.
7) Switch off the DNS-323 and remove the "other disk" - which ever one you did not remove in step #1
8) Switch on the DNS-323 and verify availability of data.
9) Repeat steps 3, 4 & 5 to return the unit to full RAID functionality

Please be aware that RAID1 is not a replacement for a proper backup regime.

Once your data is available in step #8 that should satisfy you that the rebuild was properly done - if I recall correctly the unit also displays the status of the array on the status page.

Last edited by fordem (2007-10-09 20:26:12)

I'm not doing it twice to test both disks - I'm rebuilding the "failed" disk and then removing the remaining one to prove to you that the rebuild was successful - the data being available with both disks installed (step #6) proves nothing, the system could be reading from the disk that has not been removed.

Your test procedure never verifies that the data continued to be available, making it a completely invalid test.

You can use the disk you removed as a replacement - providing you are willing to "clean" it first - if you install a disk with data, the results of the reformat/rebuild process may or may not be what you expect - there are posts on this forum indicating that both disks were formatted one by one, resulting in lost data.  I have not personally experienced this, possibly because I make a point of only using "clean" disks as replacements - as they say "experience teacheth wisdom" - many years ago, when Novell NetWare was the big name in networking, I was playing with disk mirroring in a lab, and my replacement disk had data on it and Novell very happily rebuilt the mirror from the wrong disk.

What you are suggesting will not work because - like I said - any attempt to format disk #2 with disk #1 removed (or vice versa) will result in the RAID settings being changed - the unit will not allow you to reformat the disks without changing the RAID settings, and will not allow you to configure RAID with only a single disk installed - feel free to try this for yourself.

If you have data you're willing to risk losing, just try it the way I suggested - rather than making an unnecessary purchase.  The disks I have in mine are 250GB Maxtors that were purchased over a year ago - they spent the first 6 months of their life in an IBM xSeries 206 in which they were spinning 24/7 and the last 8 months in the DNS-323 - and show no signs of distress.