ftp222

Thank you Joe L. That seems to have gotten everything in order. Running new permissions script now.

This has actually changed for ESXi 4.1. ESXi 4.0 supported only 2 devices for passthrough, but ESXi 4.1 supports 4! http://kb.vmware.com/kb/1010789 fade23: This build is fantastic, thank you for posting it. You have inspired me to shell out a bit of money and move away from my software based VMware solution into a full blown ESXi config. I will be going with the IBM M1015 cards that do support 3TB+ drives - only have 15 drives right now. I'm curious to see the new feature set of ESXi 5.0, but with the new licensing model I'm not sure many folks will be moving to it.

I just upgraded my test system to 5.0b10 from 4.7. I followed the instructions in the release notes for the upgrade. None of the drives showed MBR error or any messages pertaining to an MBR problem; however when I started the array, disks 1 & 2 show as unformatted. All disks worked fine in 4.7 for the past month. When I upgraded to 4.7 from 4.6 I did have HPA on these 2 drives. I removed it with a tool suggested by this forum and it allowed an upgrade to 4.7 without any problems. I have to believe this is somehow pertinent to what I am seeing with this upgrade to 5.0. My system log is attached. It looks like there is a problem with the superblock on md1 & md2? What steps do the experts suggest? Jul 27 22:15:44 Backup logger: mount: wrong fs type, bad option, bad superblock on /dev/md2, Jul 27 22:15:44 Backup logger: missing codepage or helper program, or other error Jul 27 22:15:44 Backup logger: In some cases useful info is found in syslog - try Jul 27 22:15:44 Backup logger: dmesg | tail or so Jul 27 22:15:44 Backup logger: Jul 27 22:15:44 Backup emhttp: _shcmd: shcmd (115): exit status: 32 Jul 27 22:15:44 Backup emhttp: disk2 mount error: 32 Jul 27 22:15:44 Backup emhttp: shcmd (124): rmdir /mnt/disk2 Jul 27 22:15:44 Backup kernel: REISERFS (device md3): Using r5 hash to sort names Jul 27 22:15:44 Backup kernel: REISERFS (device md4): Using r5 hash to sort names Jul 27 22:15:44 Backup kernel: REISERFS (device md7): Using r5 hash to sort names Jul 27 22:15:44 Backup logger: mount: wrong fs type, bad option, bad superblock on /dev/md1, Jul 27 22:15:44 Backup logger: missing codepage or helper program, or other error Jul 27 22:15:44 Backup logger: In some cases useful info is found in syslog - try Jul 27 22:15:44 Backup logger: dmesg | tail or so Jul 27 22:15:44 Backup logger: Jul 27 22:15:44 Backup emhttp: _shcmd: shcmd (120): exit status: 32 Jul 27 22:15:44 Backup emhttp: disk1 mount error: 32 Jul 27 22:15:44 Backup emhttp: shcmd (125): rmdir /mnt/disk1 Jul 27 22:15:44 Backup kernel: REISERFS (device md8): Using r5 hash to sort names Jul 27 22:15:44 Backup kernel: REISERFS (device md6): Using r5 hash to sort names Jul 27 22:15:44 Backup kernel: REISERFS (device md5): Using r5 hash to sort names Edit - I assumed I would see MBR unknown on the main page somewhere, I did not know I had to click on the disk number to see the info. Upon clicking the disk number, I do see: Partition 1 size: 488386552 KB (K=1024) Partition format: MBR: 4K-aligned File sytem type: unknown syslog07272011.txt

Thank you for keeping up with these tool packages, they are much appreciated!

Great, thank you Joe. Can this be run prior to upgrading to the 5.0 series? I suppose a follow-up question is, if this is run prior to a 5.0 upgrade and files continue to be written to a 4.7 system, will the script have to be run again (specifically against those new files)?

Still looking for an answer to my questions in the original thread. I'm hoping an expert can chime in. I just upgraded my backup system to 4.7 and cleared 2 HPA drive errors I knew I would have, so I am ready to start testing the 5.0b7 series.

I understand this may go quickly for some users, but not others. Does anyone know an answer to my disk-by-disk question and if it can be done before upgrading to 5.0?

Is it possible to run the New Permissions utility prior to upgrading to the 5.0 beta? I am currently on 4.6, but will move to 4.7 in the next few days before moving to 5.0. Ideally I would like to run the New Permissions Utility one drive at a time as some of the drives have a lot of files and I'm sure it will take a while to run. i.e. run it on one drive each night until I get through all of them. Is this possible? Is there a command or script I can use to run this or does it need to be done through the 5.0 web interface?

I am not too familiar with licensing and Linux/GPL/etc. but can't someone just post a compiled package? I know with VMware there were restrictions due to serial number and agreements for downloading the software. There is no serial number for VirtualBox, but is there another agreement/restriction? I was able to create the package without any issues, but I'm sure this would be more widely adopted if it were a simple unmenu add-on.

I don't know of a way to output the format like that and I'm not sure it is possible. I know you can use a /V to show skipped files, but I still think it displays the full path for every file, which I know can get unreadable for deeper folders. I'm not on my box right now, but you could try the /ETA switch, but I doubt that cuts off the path's. Beyond that I think you're out of luck.

I believe you need to use the /FFT switch for robocopy. I had this issue initially as well with the Linux/Windows timestamps being slightly different. I also use the /COPYALL to preserve file and directory timestamps.

This is very intriguing! Thank you for the detailed guide and I love that it can be done without a Dev system - that's a huge plus. Question to Virtualbox users - How does the speed/performance compare to VMware Server 2.0? My biggest issue with VMware is a perceived lack of speed or sluggishness at times. I may have to give this a shot and compare it myself. Joe L. if you can get an all-in-one installer going through unMenu that would be fantastic! I see a lot more people using unRAID with an easy to use VM feature.

If you are seriously considering Drobo, make sure you test multiple writes/reads at the same time if that is something you intend to do on a regular basis. I have not tested the FS model, so maybe they are better, but I owned both the Drobo and Drobo S and performance went through the floor when you tried to read/write to the unit at the same time. Also, be sure to visit the drobospace forums - are you still required to have a serial number to view the forums? Any company that will not open their forums for pre-sales support should not be a strong consideration. Don't get me wrong, I absolutely loved the concept of Drobo and thought it would be the end-all of storage, but that was not the case when i used them. Between SERIOUS performance issues and near loss of all of my data (among several horror stores on the Drobo forums), I won't go near another one again. Good luck in your hunt for a viable solution!

Well the test with my backup server didn't go quite so well. After numerous issues with a CPU upgrade, I found out VMware ESXi requires 2 GB of memory - my backup system only has 1 GB. So, this project is on hold until I can scrounge up so more memory since I will not jeopardize my production system until i can verify everything is working. Still interested to hear anyone else's experiences with this.

I have 2 unRAID systems at home - The main production system has a Q9400 (quad core) processor and 6GB of memory. I currently run VMware Server 2.0.2 (software) and 2 Windows XP VM's on the system without any issues. The main limitation I ran into for speed with this system was the slow hard drive I was using to store the VM's. I've recently switched to an SSD and it is absolutely flying now. To be honest, I could do everything with a dual core CPU, the quad is overkill and not needed. My 2nd system is my backup system of production and also used for testing (both new software and hardware). It is a Pentium D (dual core - very old) with just 1 GB of memory. The plan is to put ESXi on it and map the existing SATA drives to an unRAID VM to see if everything works and is fully functional without any errors. unRAID only requires 512 MB of memory and XP can run off of 256-512 without too many issues, so it should work for testing. With ESXi, you MUST have a 64-bit capable processor and I would recommend at least 2 GB of memory and a reasonably fast hard drive to store the VM's. Once you start running on the bare metal hypervisor versus the software one, thing speed up considerably and opens many more options. I hope others are testing this and report back their findings. It sounds like everyone has the same desire to eliminate systems that are wasting energy and lower costs. I have this vision of a NAS box being able to perform several functions and it seems like unRAID in conjunction with VMware is the closest to accomplishing it at this time. Of course if Windows Home Server had a better form of data protection (ala unraid) instead of just mirroring, then it's a whole new ballgame.

I had planned to test and verify parity from an already existing unRAID array this weekend, but had issues upgrading my CPU to a 64-bit compatible one (ESXi requires this). I will try it again next weekend and report back. On a side note, you do not have to create the vmdk image file to boot from, which is probably very confusing and time consuming to many people. You can use your already working unRAID thumb drive, but you have to tell ESXi how to boot from it by using a CD Boot Image that can boot to USB. I used a suggestion from another thread and created a CD Boot image in a few minutes with PLOP (http://www.plop.at/en/bootmanagerdl.html). With this method, you just mount the ISO image as the VM CD-ROM drive and it boots off of your USB drive instantly.

Thanks for the tip, I had not run across this in my searches! I have tried this with some SCSI drives on our ESX boxes at work, but couldn't get those to work. Evidently this may just be for SATA, so I'll give it a shot at home when I get some time.

Thanks Nyago123. I got the ISO file created and booted to the USB drive without any issues (although at USB 1.1 speed). This would work absolutely perfect if I could do Raw Device Mapping to the Drives I have started testing Hyper-V and I have the opposite problem where I can't map a USB device, but I can map Raw Disks. I guess the next thing to test is an install of unRAID to a physical drive, but I'm not sure how the licensing works. If VMware ever supports Raw Device Mapping for drives not over a SAN then we'll have this one working, but until then I think I've hit a wall.

You are correct that the CPU performance may not be huge, but a 10-20% gain can be a good boost. WeboTech is right that memory access will be the big benefactor. I didn't mean to jump into the thread and demand a 64-bit version, I just think this is where software is headed and if it is an easy re-compile and post a 64-bit version, then great lets start down that path too. As I understand unRAID 5.0 is the next evolutionary step to the software and getting the hooks into it is the most important. But once we have those hooks, it would be great to take advantage of everything we can throw at it. unRAID has been instrumental in the way it handles and protects our data and nothing else out there exists that accomplishes the task in the same way with the same confidence; however other devices are catching up quickly and I just want to see unRAID do everything else the major competitors are. 5.0 will be a massive leap forward, but with memory getting cheaper and bigger, 64-bit is the next logical step so why not start testing and working out the bugs now if it is easy.

64-bit will have no appreciable effect on speed, except in rare situations. I am not looking to increase the speed of unRAID, only the other applications running on top of it. The main reason for unRAID 5.0 is to make it easier to install and have other applications run on top of it, so I believe support for a 64-bit environment should be near the top of the list to help performance of these other applications. I understand this may not initially seem like a big deal, but there are many people running a myriad of CPU intensive applications (PAR2, ZIP/RAR, VMware of course, etc.). In a Windows environment, it is pretty simple to compile for 64-bit, I'm just hoping it's similar for Linux.

How about a 64-bit version of unRAID? Can we start testing this with the 5.0beta series? I'm sure many of us would like to start running x64 apps on our processors to speed up CPU intensive applications.

What did you use to create the boot time ISO to allow boot from USB? Is it possible for you to post the ISO here? I think that may be a very helpful utility for many people to use when testing out ESXi.

That would be helpful... We can then react and more forcibly terminate a process holding a disk busy, if that is the desire of the user. I agree with this request. It would also be extremely helpful to be able to issue additional commands when the stop button is pressed. For example, I run VMware and need to manually issue a "/etc/rc.d/init.d/vmware stop" command every time I stop the array since VMware is running.

I was trying something similar last week and noticed the USB boot option was not there. Thanks for the ISO suggestion, I never would have thought of that. Does the Flash drive still show up and is writable? How does it handle Lime Tech's key file? This is high on my list to try as I am getting a bit tired of the slow performance of VMware Server software. In the mean-time I am trying to get things working with Hyper-V (MS is soooooo far behind VMware in their configuration options).

This thread got me really excited about the prospects of running ESXi and then having an unRAID and Windows VM run on top of the VMware hypervisor.... so i did a bit more research and testing and came to the following conclusions. 1. ESXi 4.1 now supports USB device passthrough (something that the enterprise community has wanted for ages mainly for USB dongle and key support). So the theory is pop the unRAID USB stick in, create a VM with the USB stick, and boot directly too unRAID via USB as we always have. Unfortunately I cannot find a way to make this work. I got the USB (unRAID) device mapped fine, but when you go to boot to it in a VM, there does not appear to be an option to boot to USB in the Virtual Machine BIOS If anyone knows how to get this to work, I'm all ears. The only option I see here is to put unRAID on a drive and boot it that way - I know I saw a thread somewhere that shows how to do this. 2. I was hoping to use the Raw Disk Mapping (RDM) feature of ESXi to simply upgrade my existing unRAID server, then map those dives to a VM and have everything pick up as I left off. Unfortunately this is not the way RDM works. You must have an iSCSI or Fibre Channel SAN target to send RDM command through. Strike 2, bummer I believe Microsoft's Hyper-V might be an option for this though as I have seen several Windows Home Servers being run this way... I'll have to research further, but I know Hyper-V does not work for USB booting either. 3. The only way I can see ESXi being used today is with large (disk size) Virtual Disk files. This would certainly be interesting and probably work fine, but I'm not overly comfortable with it. Any tiny bit of corruption could render the whole file useless and lead to lost data. I had high hopes this morning, but most have been dashed. I would be very interested to hear from anyone that has tried using ESXi and what your successes were like. In the mean-time I'll continue to research Hyper-V and see if that is even an option.