Adding a brand new drive (fresh from Best Buy). I have a SuperMicro case and board that supposedly allows hotswap, but this didn't go so well.  I neglected to stop the array, thinking that it wouldn't matter since this would be an unassigned device at first.  Inserted the drive without powering down, GUI didn't see it after refreshing.  Restarted via the GUI, and watched it boot normally to the Tower Login prompt.  However, I cannot reach the GUI anymore nor the VM that autostarts (via Chrome Remote desktop on another laptop).  Seems like it's there because I can ping the IP, and the login prompt is there, but no GUI.  Doesn't matter whether the new drive is physically connected or not.

 

I have nothing of consequence on it yet, but what did I manage to do here? Is it really this easy to bring everything down?

Just realized that I had not connected an sata cable from mobo to the backplane, which would explain why unraid didn't see it. But now I am even more concerned about the stability as I have no idea why any of this would affect the GUI. I can reach the mobo via IPMI and I can ping the tower ip. No GUI and no shares accessible from another computer

Login on the console and type    diagnostics   at the command line prompt.  That will write the diagnostics file to your flash drive in the   logs   directory/folder.  Upload that file with your next post. 

 

I have a note here that you can type  poweroff   to shut the server down.  (I have not tried that yet...)

 

EDIT:  I know that hotswap may work but I am old-school and I would never try it.  I believe that it requires hardware, BIOS and OS support to work properly. 

Edited August 21, 20178 yr by Frank1940

Thanks for responding. A few more reboots and it was up and running again, this time with the SATA cable attached, so the drive is now preclearing. It seems no worse for the wear except that the preclear plugin was gone and had to be reinstalled. Also, the bjp script was no longer there.  No idea on that one.

 

Before it magically started working again, I used putty to ssh in and use the logistics command from the CLI, then I shutdown and put the flash in another computer to send it, but I couldn't find it (maybe hidden file? did not think of that until this morning).  Put it back in and rebooted to see if I could get at it another way, that's when it came back online.  I ran the attached diag file from the GUI AFTER it was up again, so I don't know how useful it is at this point. I assumed since they are time stamped the first one would be there somewhere, will look again tonight.

 

Now I'm thinking about whether my "Staples Relay" branded 4gb sandisk flash drive is to blame.

 

Thanks all!

 

 

tower-diagnostics-20170820-2351.zip

3 minutes ago, btrcp2000 said:

I used putty to ssh in and use the logistics command from the CLI

 

To get the diagnostics file (like the one you uploaded), you type diagnostics on the CLI. 

 

It is best to avoid the use of third party house brands for your boot drive.  You can't be sure of the quality.  (An even worst choice is one of those 'free' flash drives that are often given away as advertising promotions.  Google "rubber ducky usb" for the reason.) 

 

It is also best not to use a USB3 port (they have a blue plastic insert) for your unRAID boot drive.   Sometimes, the boot drive ends getting 'lost'.

I did type diagnostics, the above was an autocorrect. I know I saw the CLI put out a timestamped file name after the first attempt, but it was late and I was fading. Will dig some more tonight. Would the second one have overwritten the first?

 

Motherboard is SuperMicro x9dri-lnf+, which doesn't have USB 3.0. I am using the 2.0 port that is directly attached to the surface, not the I/O on the back.

 

Will do some research on flash drives and put something better in.  I remember having that concern when I first setup a "demo" unraid server a few months ago, but forgot about it when everything popped right up.

18 minutes ago, btrcp2000 said:

Would the second one have overwritten the first?

What you get  on the CLI is this:
 

diagnostics
Starting diagnostics collection... done.
ZIP file '/boot/logs/rose-diagnostics-20170821-1104.zip' created.

As you can see, it should have a unique file name because of the date and time string.  (There is every reason to believe that the GUI actually runs this command with some sort of switch to download it to your browser rather than saving it in the logs folder.)  If the file wasn't/isn't in the logs folder, something was wrong with either the flash drive or its mounting to the Linux file system.  That is why I cautioned you about the UBS3 port...

Edited August 21, 20178 yr by Frank1940