I'm flying blind...

[TapRackPull]

So I started a copy of data from a 1TB dive into my 3x2TB RaidZ1 over 30hrs ago. Best guess it was a scad over 700GB of data. Now, if I put my ear near the server I can hear random ticks of what I hope are the heads of the array engaging... Maybe its just hope.

I have 4GB of RAM...

I absolutely, under no circumstances can login to the GUI or SSH into the server.

I intend to wait for 48hours as that would allow for a transfer rate of nearly 4MB/s. After that, I will shut it down - whatever that takes. And start over. With rsync this time...

The thrust of my question is whether or not increasing to 8GB of RAM allow me to login despite my other actions. I NEVER want to deal with the 'not knowing' side of server BS again.

 

[ProtoSD]

You can't get to a shell at the console?

Are you running AMD64 or i386 version?

I'm guessing your vm_kmem values may need to be adjusted because I had a similar problem in the beginning, but 30 hours?!! You can probably type 1 & 0's faster than that! ;)

 

[TapRackPull]

I am running AMD 64 Beta 4.

Protosd I have no prob "pulling the plug" on my current action deleting the volume and creating a new volume if I know there is a a happy ending.

Having said that, I have NO idea how to 'adjust' vm_kmem or to tell it what to adjust. Although I can do it... if, you tell me how. I just wouldn't presume to know what needs 'adjusting'. Although, my first guess would be a lot.

 

[ProtoSD]

If you did pull the plug or just hit restart, you wouldn't loose your pool or any data that copied already. This is a good example of why using rsync is useful. It can figure out where your copy left off and leave out the guess work.

I could guide you to adjusting those settings, but it would be nice to see some type of error message on the console that might give a clue to what the problem is before we just go with my hunch.

So, do you see any errors on the console or can you get to a shell on the console?

 

[TapRackPull]

On the consoe there is NOTHING since I started the transfer. The last message I have on the console is ~30 minutes pre copy command... and is in regards to an inability to open a socket. I am going to power down the system and see if I can login to the GUI, I will post back when that is done.

 

[ProtoSD]

Hmm, wish you could have tried to get to the command line on the console, you could have looked at the log files to see if there was anything useful. Not being able to open a socket? before you started to copy? that sounds like a potential problem too.

 

[TapRackPull]

This is ridiculous,
After powering down and then up, I am getting the following errors repeatedly.
After reaching the Console setup menu...
ATA 5: timeout waiting for write: DRQ
ATA 6: timeout waiting for write: DRQ
ATA 7: timeout waiting for write: DRQ

I will power-down, remove the USB, format, reinstall 8 AMD 64 Beta 4 and try again.

Will re-post

 

[survive]

Hi TapRackPull,

Can you tell us what your motherboard or controller is?

-Will

 

[TapRackPull]

Will
Here are my system specs:
MOBO: ASUS M2N32-SLI Deluxe
CPU: Phenom Quad-core 9600 2.6 GHz
MEM: Kingston 2x2GB (4 total)
PSU: Corsair 500W
HDD: 3x 2TB WD EARS
1x 1TB WD Caviar Black

I do not have a SATA/SAS controller. Everything is on the MOBO

 

[TapRackPull]

Unbelievably, I can not even get the system to reboot of the USB now.... AND I can't seem to get any version of the distro to boot from CD to create a new bootable USB. I'm pretty sure God is mad at me.

So, I will post back when/if I get anything up and running and we can take it from there.

 

[globus999]

I am running AMD 64 Beta 4.

Protosd I have no prob "pulling the plug" on my current action deleting the volume and creating a new volume if I know there is a a happy ending.

Having said that, I have NO idea how to 'adjust' vm_kmem or to tell it what to adjust. Although I can do it... if, you tell me how. I just wouldn't presume to know what needs 'adjusting'. Although, my first guess would be a lot.

There is a ticket already opened for this one.... I guess this is not a priority for the project team... go figure....

 

[globus999]

S

I have 4GB of RAM...

The thrust of my question is whether or not increasing to 8GB of RAM allow me to login despite my other actions. I NEVER want to deal with the 'not knowing' side of server BS again.

I "feel you"... In reply to your answer, no. Going from 4 to 8 will make absolutely no impact on your ability to interact with the system. There are people "out there" that are running 8.0B4 on 2 Gb RAM quite happily. You must search for the problem somewhere else.

 

[globus999]

This is ridiculous,
After powering down and then up, I am getting the following errors repeatedly.
After reaching the Console setup menu...
ATA 5: timeout waiting for write: DRQ
ATA 6: timeout waiting for write: DRQ
ATA 7: timeout waiting for write: DRQ

I will power-down, remove the USB, format, reinstall 8 AMD 64 Beta 4 and try again.

Will re-post

There is another ticket opened for this one. It is a flawed implementation in nanoBSD - ignore the errors

 

[TapRackPull]

So just to further complicate this... Now, everytime I attempt to boot into fN I get to passed the countdown menu screen and it continues to boot up to the point in the picture. I cannot get freeNAS to finish the boot.



Any ideas? I am getting close to just throwing in the towel entirely. Followed by my computer in the nearest lake...

 

[Darkaine]

I "feel you"... In reply to your answer, no. Going from 4 to 8 will make absolutely no impact on your ability to interact with the system. There are people "out there" that are running 8.0B4 on 2 Gb RAM quite happily. You must search for the problem somewhere else.

Actually mine has only 512mb and while its close to its limit, on UFS it does the job. I may however upgrade to 1GB at some point.

 

[TapRackPull]

To anyone still paying attention to the ramblings of a fool.

I have successfully rebooted into 8-Stable, recreated my volumes, re-mounted my ntfs drive and would like to copy over data. This time with rsync.

I have no intention of doing so until someone can point me in the right direction. I would rather not do this again.

Thank you everyone for you time and help with this. The end is in sight!
Matt

 

[ProtoSD]

TapRack.... Glad you got things going again. Did you figure out what went wrong so you don't get bitten again?

Here's a thread about Rsync:

Best-way-of-backing-up-a-FreeNAS-server-to-another-FreeNAS-server&highlight=rsync

 

[TapRackPull]

Nearest I can tell, the transfer was going just fine. It seems that everytime I try to copy large quantities of data my ability to connect to the server drops out. So this time I will do it via the shell and not SSH so I won't become blind instatnly. Seems wrong all the way around.

Before I do this are there any settings that I should make sure are taken care of to maximize the tgransfer rate? AIO? large RW support?
Is there a way to check my write speed or should I just plan on 7-9MB/s???

 

[TapRackPull]

Proto bearing in mind that this is local to local transfer would the following be an ok command
rsync --partial --progress --stats -av /mnt/ntfs/Movies /mnt/Media/Movies
Is my leaving off -r or -z an isssue?
Also, if I wanted to access a directory with a space in the name for example "My Folder" how would I write that since if I type a space fN thinks I'm done with that argument?

 

[ProtoSD]

Man, it's getting tough for me to find all the threads with stuff to point people to! I *think* I saw it was better with AIO disabled and Large RW enabled, could be wrong. 7-9MB/s is awful... but that could be your router/switch too. Your connection to your server dropping sounds like a problem. I would just give a try to adjusting some settings in your loader.conf file. Here's what you do:

1) login from the console (#9 on the menu)
2) su (enter the password for root, should be 'freenas')
3) mount -uw /
4) cd /boot
5) edit loader.conf (use whatever editor you know on Unix... 'vi', might be your only choice)
6) insert/add the following lines anywhere in there that pleases you & save the file

Code:

vm.kmem_size="1536M"
vm.kmem_size_max="2048M"
vfs.zfs.arc_min="256M"
vfs.zfs.arc_max="1024M"
vfs.zfs.prefetch_disable="0"

7) reboot your system with 'shutdown -r now'

These are the settings I use with 4GB, if you make any typing mistakes, you may not be able to boot. Then you might end up re-flashing... again...

If you can boot back up, then I think you're ready to try copying again.