Moving data on new disks

[NeverSimple]

Hello, I'm new to Amahi, coming from Windows Home Server. My WHS died 2 weeks ago and I really didn't want to buy yet another OEM license because i had to replace hardware (motherboard). So i went looking for a replacement. What got me interested in Amahi was the Greyhole storage pool which i think is a really great idea. Thanks for developing such a thing!.

Aaaanyway...

1- I'm testing Amahi/Greyhole with a 160GB system drive / landing zone for the shares. After basic Amahi install i added two 250GB disks and added them to the storage pool. I've set up most of my shares to use 1 extra copy but the largest, "movies" to use 0 copies. The 2 drives are are filling up quick and I'll probably replace them with the 1TB drives from my WHS if i decide to continue after the test. I've read about the "--going" command, but I'm pretty sure one of the drives can't hold all the files, because of the large "movie" share which is now split between the 2 drives. What would be the best way to handle this? (replace BOTH smaller drives with 2 1TB drives). It took a long time to copy these files from another machine and I don't want to mess up things because I'm doing something wrong....

2- I'm analyzing the Greyhole logs, to see what it is doing. I think I understand the reason for the " 7 sleep: Nothing to do... Sleeping" message every 10 seconds (easy to see if the daemon is still running, I assume), but it fills up the log very quickly. Wouldn't it be possible to just add this 'sleep' message once and only add another message to the log when it starts doing something else? I guess you could check on the daemon with an extra command line argument like "greyhole --status" or something, that could also go into the log " 7 status: sleeping for 2400s" or something. Not that important, just a thought.

Again, I'm impressed by Greyhole; it seems to be unique in the Linux world. I'll hope that Amahi / Greyhole will make some real competition for WHS.

Richard

[gboudreau]

1. The best way to handle this would be to add another HDD, before you remove any of your old drives.
--going will then have enough room in the pool to move the files off your old drives.

If you don't have any IDE/SATA ports left, you can connect the new drive using USB, then execute --going, then remove the old drive, and connect the new drive on the internal port instead.

If you can't do that for some reason, you can still run --going, which will fill your other drive to capacity, then remove the old drive, and replace it with a new bigger one. At that point, connect the old drive anywhere (on your server, or any client), and manually copy your missing files from your old drive 'gh' folder into the new drive 'gh' folder (/var/hda/files/sdb1/gh/* for example).
The next fsck will find those, and the files will re-appear on your shares at that point.

2. The log is currently using DEBUG level, because Greyhole is actively developed. By changing the log level to INFO instead of DEBUG, the log will be much smaller, and the sleep message would only appear once every 10 minutes (I think, that's without looking at the code...)

[NeverSimple]

1. ... If you don't have any IDE/SATA ports left ...

Mhh, you must be clairvoyant or something.... I forgot to mention that

But, it seems that my WHS is REALLY dead, and in that case i'll buy new hardware with more SATA ports, for my new Amahi server. (the testserver is using a LOT of energy; old P4). So the scenario will change. I'll probably come back here and ask how to mount two existing 'greyhole disks' in a new server, but maybe i'll know more about it by then....

2. The log is currently using DEBUG level ...

That sounds very logical; could have thought of that myself.

Thank's for now.

Richard