February 1, 201214 yr I am currently running a 5b11 build using 2x M1015 and 1 MV8 cards and 9 2 TB drives. Currently, they are all on one M1015 card. Should I move 4 of them over to the 2nd M1015 and 1 over to MV8 or should I just fill up, one card at time starting with the M1015 and then MV8?
February 2, 201214 yr So say you have the ESXi machine with unRAID as a vm with 10 mapped drives. You will also have Windows 7 installed, a vm, of course. Can the Windows 7 vm touch the data on those 10 drives? Read only, read/write, or anything? If not can it access the drives faster through the machine I/O some how instead of LAN and mounting the drives with SMB? What I am getting at is being able to have an unRAID machine/vm with Windows being able to access the bits at fast speeds as if it was native to the Windows machine? Say cut and paste from one unRAID drive to another but by using Windows. Thanks
February 2, 201214 yr So say you have the ESXi machine with unRAID as a vm with 10 mapped drives. You will also have Windows 7 installed, a vm, of course. Can the Windows 7 vm touch the data on those 10 drives? Read only, read/write, or anything? If not can it access the drives faster through the machine I/O some how instead of LAN and mounting the drives with SMB? What I am getting at is being able to have an unRAID machine/vm with Windows being able to access the bits at fast speeds as if it was native to the Windows machine? Say cut and paste from one unRAID drive to another but by using Windows. Thanks If you mean RDM when you say "10 mapped drives", Windows7 is not going to read them, or write them. Nine of the ten will be reiserFS, and the parity is on the last one. This thread does not use RDM. The controller is passed to the guest and unRAID alone sees the controller and connected drives.
February 2, 201214 yr So say you have the ESXi machine with unRAID as a vm with 10 mapped drives. You will also have Windows 7 installed, a vm, of course. Can the Windows 7 vm touch the data on those 10 drives? Read only, read/write, or anything? If not can it access the drives faster through the machine I/O some how instead of LAN and mounting the drives with SMB? What I am getting at is being able to have an unRAID machine/vm with Windows being able to access the bits at fast speeds as if it was native to the Windows machine? Say cut and paste from one unRAID drive to another but by using Windows. Thanks If you RDM when you say "10 mapped drives", Windows7 is not going to read them, or write them. Nine of the ten will be reiserFS, and the parity is on the last one. This thread does not use RDM. The controller is passed to the guest and unRAID alone sees the controller and connected drives. I think I understand what you are explaining. unRAID is the only OS that sees the that sees the controller and no other OS can see it at the same time. If that is correct then is there something that can be ran in unRAID, software or a vm, that can see everything? "See" meaning view, read, and write the data on there shares.
February 3, 201214 yr I think I understand what you are explaining. unRAID is the only OS that sees the that sees the controller and no other OS can see it at the same time. If that is correct then is there something that can be ran in unRAID, software or a vm, that can see everything? "See" meaning view, read, and write the data on there shares. Why not just use SMB to access space on unRAID from other machines?
February 3, 201214 yr I think I understand what you are explaining. unRAID is the only OS that sees the that sees the controller and no other OS can see it at the same time. If that is correct then is there something that can be ran in unRAID, software or a vm, that can see everything? "See" meaning view, read, and write the data on there shares. Why not just use SMB to access space on unRAID from other machines? Because having to move files around on the "Tower" using SMB would need the files to transfer from "Tower" to SMB viewing machine and back to "Tower". I could use telnet but that is no fun and not easy at all. It's geeky, which is cool, but it takes way to much time to move around tons of stuff. I am looking for a GUI to view and transfer bits on the "Tower" but actually at the machine itself. Like telnet but with a GUI. Does anything exist like that? Thanks
February 3, 201214 yr type "mc" once you have a telnet connection to the machine. Wow, super quick reply. That is interesting but bot close enough to what I am looking for. Thanks prostuff1.
February 3, 201214 yr You can also use Midnight Commander ("mc") to select multiple folders rather than doing it one file at time. Figured tit out after moving 100+ folders. Doh! Chec out this tutorial: http://www.trembath.co.za/mctutorial.html
February 4, 201214 yr Because having to move files around on the "Tower" using SMB would need the files to transfer from "Tower" to SMB viewing machine and back to "Tower". Before you start making stuff up, you should try it. If you open two explorer windows to "Tower" and drag a selection between them, Windows asks if you want to move and the data is not transferred over the network. You might not want to post on a hardware design thread if you want help with client interaction. Even add it to right-click menu
February 4, 201214 yr JohnM (& others), How many CPUs should i assign to unRaid in my VM guest? Thanks for this great guide, I got my 2nd unRaid box up and running with the SCM-F-0, E-1240, 16GB RAM, & a M1015 card.
February 4, 201214 yr How many CPUs should i assign to unRaid in my VM guest? As few as possible because adding more then is needed can cause it to be slower. One is working fine for me. This is due to CPU resource scheduling, if you have 2 CPU's it will wait for 2 physical CPU cycles to open up to run the job, if you have 1 CPU then you only have to wait for 1 open CPU cycle. The typical wait time for 1 open CPU cycle is typically shorter then for 2 concurrent open CPU cycles and therefore the two CPU VM will have to wait longer for the CPU resource which in the end-user perspective makes it seem slower. ESXi 5.0 improves this a lot, but it is still best practice to use as few CPU's as possible.
February 4, 201214 yr So this is obviously a limitation (or shortcoming) of VMWare and has nothing to do with how unRaid handles multiple processors, correct?
February 4, 201214 yr So this is obviously a limitation (or shortcoming) of VMWare and has nothing to do with how unRaid handles multiple processors, correct? Correct. Its debatable if this particular shortcomming would even be noticed in a home user setting.
February 5, 201214 yr Author Hey guys, Sorry i have not even been on the internet for over a week. major work project out of town... Hey Johnm, I just followed your guide for setting up apcupsd and (without testing the shutdown) things appear to be set up correctly. However, I keep getting COMMLOST errors followed by ONLINE pretty consistently even if I up the polling time. Considering unRAID and the Windows VM are on the same host, I would assume that communication errors shouldn't be an issue. Did you ever experience anything like this before? Thanks!! I see this once in a while between my 2 unraids. maybe once every 2 days and it lasts all of 1 second.... you might try going to the older version of ACPCUPSD? Other then the few bits of spam in my email, I am pretty sure it wont effect the script if you had a power failure. I never bothered looking into it. if you see the fix, let me know.
February 5, 201214 yr Author JohnM (& others), How many CPUs should i assign to unRaid in my VM guest? Thanks for this great guide, I got my 2nd unRaid box up and running with the SCM-F-0, E-1240, 16GB RAM, & a M1015 card. Just one unless you are running add-ons in unRAID that will need CPU like some sort of transcoding application... By default, unRAID only utilizes a single CPU.
February 5, 201214 yr Author ugh @ quad post.. I just wanted to point to those following this thread. I have still not had a single redball since migrating to the expander. I never determined what was causing it in the first place. but, I'm happy. The SataIII SSD is a nice addition for a cache drive. it has been worth the cost. The unraid server is no longer the bottleneck in my transfers.
February 5, 201214 yr PS. in testing my scripts... i found it easier to plug everything that was plugged into the ups into the wall, then when i pulled my UPS power, there was no load on it so i could keep testing real quick without having to wait for the battery to refresh. Once i had it all working. I put everything back into the UPS and did it for real. the best part is. The next day, the power DID go out and it all worked. I am going to look into getting the UPS to auto power off sometime this week. that part failed me. Great news that powering off ESXi can be done with an APC Backups instead of needing a SmartUPS with a network card installed but I was wondering about the part that I bolded. JohnM or any one else, is there a way to have the APC unit shut off as well or it just has to die a slow death when the battery runs out? Thanks for the write up though on how to make it shutdown safely!
February 5, 201214 yr Author APCUPSD wants to shutdown the UPS by default after it shuts down the box that is plugged in VIA USB. I manually disabled that. The problem is.... I still have my ESXi server on after my unraid powers off (the kickoff point)... if unraid turns off the UPS... MY Second unraid server, my windows server (with a hardware raid) plus my ESXi box and all other guests that are still in the process of shutting down would die a horrid "power off" death.. While letting the ups drain itself is bad... powering off database and raid servers mid shutdown or while still stopping the databases getting ready to shutdown could be worse. Since power failures are a rare animal this time of year. i have never looked into it. but my guess is that since the USB is plugged into a box that is already off, there is no way to send a command to it to power off. in order to have it work, the PC with the usb would have to be the last box off... unless there is a command to "shut off the UPS in 15 min" command that i am unaware of... if i bought a second UPS (or tossed out 2 servers), i would not have this dilemma. obviously everyones situation is a little different.
February 6, 201214 yr APCUPSD wants to shutdown the UPS by default after it shuts down the box that is plugged in VIA USB. I manually disabled that. How did you do that? I have not played around with this. BTW, way I have set UPS in my box is I added a USB device for the unRAID VM. Instead, I think it would be better if: [*]have another guest VM, like a Windows or a linux box, with the UPS assigned to it. If using Windows VM, you can run a number of script when power-loss is detected using the included Windows software [*]this guest runs the powerdown script remotely for the unRAID VM. [*]After unRAID is off, suspend the other VMs using VMWare PowerCli -- see step 3 The only question is - is it possible to run the powerdown script remotely. I would assume so since I think unMENU has this option but I don't have a lot of experience with linux script yet. Do you think this would work?
February 6, 201214 yr Usually APCUPSD hooks itself into the shutdown file /etc/rc.d/rc.6. From the looks of this snippet, the command /etc/apcupsd/apccontrol killpower is what tells the UPS to shut down after a preconfigured time with no power. # This will cause the UPS to kill the power supply after a configurable number # of seconds (see the apcupsd.conf file). This is important for situations # where the mains power comes back before the UPS batteries are completely # dead. if [ -f /etc/apcupsd/powerfail ]; then echo echo "apcupsd will now power off the UPS" echo /etc/apcupsd/apccontrol killpower echo echo "Please ensure that the UPS has powered off before rebooting." echo "Otherwise, the UPS may cut the power during the reboot." echo halt -f -p # power down the system if the hardware is capable fi killpower apcupsd does not normally generate this event. Instead, it is invoked directly from the system halt script as 'apccontrol killpower' because the killpower event needs to be performed as late in system shutdown as possible. Default action -- sends "UPS now committed to shut down" to wall, waits for 10 seconds and then issues the command "apcupsd --killpower" to put the UPS into hibernate mode and shut off power to the connected equipment. In the case of a smart UPS, the UPS will then wait for the expiry of any configured shutdown time delay specified by the SLEEP configuration directive. In hiber- nate mode, the UPS will restore utility power to the connected equipment subject to the values specified for the RETURNCHARGE and WAKEUP configuration directives.
February 6, 201214 yr Author @ashaneil I have 3 Physical servers on that UPS. 1 unraid, 1 Windows 2k8 with a massive hardware raid and my ESXi box. It would be best if i had the windows box control it all. Unfortunately, the ESXi box is the only one guaranteed to be on 24x7. I did try to set up my windows script PC to handle the UPS at first. for some reason it was not 100% reliable. I always said I'd go back and look at it. I just don't really want to touch a server once I get home after a 12 hour day in the data center at work. it may be ghetto how i set it up, but it works every power fail. [so far....] @BRit... Yeah it looks like that might be my answer.. again.. time is something I wont have till mid summer due to work projects stacking up...
February 6, 201214 yr Author I decided to try yet another WeeboTech script on my server... This one should test my cache drive write speed.. Not quite what i expected... but not to shabby. the lack of trim might be slowing my drive down a little.
Archived
This topic is now archived and is closed to further replies.