overbyrn

I had a go at a CUPS plugin some time back but never finished it. Not at a PC right now or would dig it out. I might be able to advise the original poster as I did get AirPrint on unRAID working and have these days done the same on a Debian VM. Will re-read the first post tomorrow when not on a mobile device.

If you have time mrow, a write up of the steps involved including links to the dev kit, command line args for vmware-mount.exe etc, would probably benefit a ton of people. Maybe BetaQuasi could include it on the first post or link it to a wiki location? I guess the truth is, if you're running this type of config then the implication is you'd know how to do it, but well you know...

A small piece of PayPal advice... if you open a case with PayPal, ensure the case remain open until you reach a satisfactory conclusion to the dispute. Never ever close the case as you cannot reopen it. eg. if the seller requests the case be closed ahead of a refund, let this be a big red flag to you. Once a case is closed, you have no further recourse for a dispute with PayPal and the seller and will be left having to deal with your bank or card company to get a refund. I don't wish to suggest Seanant is anything other than an honorable person, but I must say recent comments in this thread mirror exactly what happened to me when trading with him last year (also for an M1015). Long delays in communication followed by messages indicating item shipped and then later after no item arrived, less than clear replies as to why. First felt something was wrong when I pressed for exact shipping dates, at which point I never really got clear answers to my questions. Raised a dispute with PayPal, reply from Seanant requesting case be closed and a refund would then take place. That was second time something felt wrong, which is when I checked the dispute rules and found closing case would essentially leave me out in the cold if a refund didn't take place. Informed Seanant I would not close case and to be fair he immediately performed a refund. Happy, I closed the case and thought nothing more of it. Months past and then quite oddly and totally out of the blue, I received a personal message from him via this forum whereby he repeated his stance that I was the only person he'd ever had an issue with. I genuinely hope Seanant is on the level. I imagine he's successfully sold to others on here, so perhaps it's just a case that so many people want M1015's that he struggles to fulfill orders. It's just that this thread is giving me a weird sense of Deja Vu and I felt I should share my experience.

Apologies for resurrecting an oldish thread but was looking at shellinabox and thought I'd do a search on the UnRAID forum for any other chatter about it. This looks to be the last thread relating to it and I hit the same issue of some keys not working. In case anyone is interested, I've applied a patch to the source and compiled a fresh package. It seems to work for me. Thought I'd share here in case anyone else came looking for the same thing. Link to version that works with FF here

Loophole, Suggest you grab the Logitech Media Server plugin ... http://lime-technology.com/forum/index.php?action=dlattach;topic=18310.0;attach=9809 This was written by Tom and presented as the first working plugin example. It's heavily commented and should give you a good starting point. Regards, overbyrn

Not strictly correct. An earlier post from Tom ... http://lime-technology.com/forum/index.php?topic=14665.msg138228#msg138228 /boot/plugins/ = "system" plugins /boot/config/plugins/ = "user" plugins

Beta, Thank you for your work on doing this. I have had an APC UPS since December last year but soon after I virtualised my UnRAID environment and have only just got around to setting this up. Followed your guide pretty much all the way, but I thought I'd post here as the only issue for me was that I don't allow root as an SSH user or password access of any kind. The only users I do have are setup with private / public keys. This caused me to have to change a few things slightly in order to still be able to use plink. I thought I'd document here in case it helps anyone else. Basically all I did different was on my Win7 VM which runs apcupsd, I first created and saved a session from Putty. In the session I made sure to point it at the private key I was using. Then my plink command for performing the shutdown looks like this; plink.exe "Tower" -l apcupsd "sudo /sbin/powerdown" In this case, "Tower" is my saved session in Putty and "apcupsd" is my SSH user (which I created in the standard way from inside the UnRAID webgui and subsequently selected in my OpenSSH plugin as a valid SSH user". You will notice that I also have sudo ahead of the powerdown command. This is the only way I could find to sucessfully call the powerdown script initially as a non-root user but eventually as sudo root in order to satisfy the permission needs of the powerdown script. This meant I needed to modify the default sudoers file on my UnRAID. I made the following two changes: # User privilege specification root ALL=(ALL) ALL apcupsd ALL=NOPASSWD: ALL # Defaults specification Defaults:apcupsd !requiretty I could be more restrictive by ensuring my user is only able to run the necessary commands, but for my purposes this is sufficient. I've added a line in my go file to copy over my custom sudoers to /etc upon reboot. Still testing, but seems to work for the moment! Kind Regards, overbyrn

I think you need this package also: http://slackware.cs.utah.edu/pub/slackware/slackware-13.1/slackware/n/ppp-2.4.5-i486-1.txz It contains the missing binary. pptpd seems to run after this is installed...

The kernel modules from previous version will not load into the kernel used in unRaid5.0rc15. The modules from the current version of open-vm-tools do not build correctly against the kernel used in unRaid5.0rc15 so it may be some time before there is a new version available. Hi Zeron, Is the vmhgfs module a showstopper? Reason I ask is that I was having a go at compiling this myself. Looks like I picked a bad time to learn eh. I notice there are two issues depending on which snapshot release is used: open-vm-tools-2012.10.14-874563 gives an error in inode.c during compile of vmhgfs open-vm-tools-2013.04.16-1098359 does compile, however due to vmci and vsock being merged upstream, it breaks the loading of vmhgfs module. I found a bunch of posts regarding patches for other distros, but afraid I'm not clever enough to work out how to adapt them to fix this problem. Regards, overbyrn

Same thing here too. Had it happen a few times recently whilst working on plugins. No decernable pattern that I've noticed. Win 7 client PC's, rc12a Unraid. Can't say I've seen it on my main Unraid. That's still running rc11.

How do I find, which add-on has the bad xml plugin file? I am only using the following plug-ins, Power Down, Monthly Parity Check, AirVideo and JRE. Do, I've got to uninstall one by one and fine out that way? Your log indicates the error is with something trying to be installed ahead of AirVideo. Assuming plugins are installed alphabetically based on their name, I'd check to see what else installplg might be trying to execute in your plugin dir. As an aside, you're using a very early version of AirVideo. You should update (jre is now included).

Regarding ways to mount a VMDK under Windows. Here's another approach: http://www.running-system.com/how-to-mount-a-vmdk-file-under-windows/

@ speeding_ant & bonienl Can I ask why it is in the plugin "simpleFeatures.email.notify-1.0.10-noarch-1.plg" you are choosing to delete /usr/include? I've just spent an hour trying to understand why both nzbget and Denyhosts plugins are no longer working after installing your plugin. In the case of DenyHosts, it's because standard python2.6.6 files reside under this location. In the case of nzbget it's because libpar and libsigc++ packages install to this location. I see no reason why this location needs to be removed via your script. Please can you advise whether this is intentional and the rationale behind it as at the moment it's killing other plugins and likely impacting on anything which relies on running "python setup.py install". EDIT: perhaps unless you have an objection, the plugin is changed with prune set to false by default. I don't think it should be up to the plugin to automatically decide for the user what directories are classed as unimportant and safe to remove. Regards, overbyrn

Can you tell me where you brought your Supermicro X9SCM-iiF motherboard from. I ordered one from Lambdatek and they held onto my money for over a month before telling me again I was going to have to wait another two weeks for it, terrible company to deal with. They advertise 100 in stock and take your money then palm you off every two weeks. I have had so many problems with items I ordered from them being faulty or broken I wouldn't touch them with a barge pole ever again, but I really want this motherboard and Im not paying £280 that Amazon are asking for it and nowhere else in the UK seems to have them. Thanks Probably not what you're going to want to hear, but I got my motherboard from Lambda-tek. I didn't get the iiF though. Only the F-O. Like you, I wanted the newer board but didn't fancy paying Amazon prices. For a while I played the watching game to see if anyone had either of the mobos + memory + Sandy/Ivy in stock. In the end I had to suck it up and accept if I wanted to upgrade anytime soon, I'd need to go for the older board. I'd used Lambda-tek before for occasional stuff but never anything big. I went with them simply because the pre-sales guy I spoke to seemed quite knowledgable including already knowing the issue with F-O board and needing to have v2.0 BIOS if buying an IvyBridge CPU. They also had the ram and CPU in stock. Their 100 in stock thing is a bit sleazy. The pre-sales guy told me that for any item on their site where it states 100, it means they will source from multiple suppliers and not to take that figure to mean there's 100 sat in a warehouse. FWIW, I appreciated the honesty. The CPU and F-O were "100 in stock" Translated; they were sourced from the Netherlands, but to be fair the order was placed on a Tuesday and I had the kit in hand on Thursday. So I can't really complain. I guess milage really does vary where UK online suppliers are concerned. For instance, I'd never use eBuyer for pretty much the same reasons. There are definite supply contraints for the newer board at this time. For the small benefits the iiF brings, I'd say unless you have a good reason to wait, go for the older one.

Answering my own question. For anyone else in the UK who's wanting to fully populate the X9SCM-F-O to 32gb and needs 4 x 8gb modules; Crucial or Mr Memory. Both have current stock of compatible memory modules. Personally I chose 2 x 16GB (2x8GB) kits from Crucial - however, DON'T buy it directly from Crucial unless you want to pay over the odds. Instead enter the Crucial part code (CT2KIT102472BD1339) into Lambda-tek instead. Crucial price - £143.99 inc VAT, Lamda-tek for same item - £112 inc VAT. Lambda-tek currently have stock also. Came next day. No affiliation to them, just wanted to share they seem pretty competitive and knowledgable. Had long chat with guy who new all about v2.0 BIOS / IvyBridge vs SandyBridge etc. and was spot on by saying the board would turn up with v2.0 BIOS already onboard. v2.0b in my case.

Question for speeding_ant... Is there any mileage in a static php build? I've dabbled in compiling my own php from time to time although never statically. Would that remove the need to install dependent packages? I guess it would bloat the php binaries. As it stands I need to review my own plugins such as OpenSSH and AirVideo which now use older packages than the ones in the new SF core and web. I know without a package manager this is a nightmare, but before updating to newer versions, it'd probably be wise from a change management point of view to collaborate with other plugin devs who's work may be impacted. Regards, overbyrn

Harpz, based on our earlier PM, I've knocked up a document to explain the process of upgrading the firmware. To be clear of a few things; 1. this is work in progress 2. this is work in progress (had to be said at least twice I think) 3. I've only ever flashed two cards so my experiences are limited. I defer to others more skilled or familiar to this process than I and if there's anything wrong in the below linked document, please tell me! 4. Although I only did a card last week so the process is still fresh in my head, I did it in a Windows environment. I realise you will probably want to do it from a USB bootable stick. I've documented both ways, but *I've NOT yet done it via the USB boot method*. I've culled the information from the various threads and I'm pretty sure this is the correct way to do it. Please do you're own checks before following the process. I'd prefer if someone like Ford Prefect could review said docco. I'll PM him after this. Important Note: the instructions are only for how to upgrade an already cross-flashed card. Link to the docco (PDF): https://dl.dropbox.com/u/572553/UnRAID/IBM%20M1015%20Firmware%20Upgrade%20Instructions.pdf There are probably tons of spelling mistakes and for that I apologise. I've not had a lot of time to review it yet and now have to go make some Papier-mâché fish (don't ask). I've a spare M1015 card and plan to re-do some of the screen shots from a system with a card installed. Hope it helps. Regards, overbyrn

jowi, I don't think there is a specific reason why. I did a card last week and I must admit that I chose to cross-flash it to P11 first. Why? Well, only because I had previously done exactly the same to another M1015 and I knew the rar file had everything I needed and P11 is/was a known working cross-flash. I personally see no reason to cross-flash directly to P15. A sFord Prefect said earlier, it really depends on if you want to shave off a few minutes in the flash/upgrade process. I highly doubt there's any specific reason to go to P11 first. That's my opinion and I'm no expert having only done a couple of cards. That being said, I've a buddy at work in our infrastructure team that does this all the time and he tells me he's cross-flashed to all sorts of firmware versions in his time, so again I really don't think there's a special need to go to P11 first. In fact, the main LSI thread also has P10 as an option for the initial cross-flash firmware which leads me to assume it matters not one bit what you go to. I suspect it's just there there is safety in numbers and no-one wants to brick their card so they choose the safe option and flash to what others have done first. Just my two cents.

I was going to reply with some general instructions also as I've just had to do this to a fresh M1015. However Ford beat me to it with a link I'd not previously seen. Maybe this is best asked in the main thread, but as it directly relates to the IBM and it's fresh in my mind, I thought I'd ask here also... I see a lot of forum posts suggesting it's best to use the version of sas2flash which comes in the same zip as the latest firmware. But that contradicts this from madburg That's taken from the __ReadMeFirst.txt in LSI SAS2008 Controllers(P11).rar from the main thread. I opted to go with madburg and used P7 as I was doing the update on a Win7 x64 system so it wasn't any hassle, but since there's only x86 and x64 versions of this file in his rar, it got me wondering what I'd do if I wanted to flash from USB / DOS. If I had, then most of the instructions say to exchange the sas2flash binary for the lastest in the firmware zip. So my question is whether his note only applies sas2flash x86 & x64 binaries and if so, what's the reason behind needing to use P7? Regards, overbyrn

Cost. My chosen supplier assures me (in writing) the X9SCM-F will come with rev 2 BIOS. I'll trust them and should it not, it'll go back.

Ok... so like many others before me, I'm going to pull the trigger on the Supermicro X9SCM-F-O mobo and an appropriate Xeon CPU. Have done my due diligence and am pretty happy with everything. But for one thing... is there anywhere even remotely UK based that is selling 8GB sticks of memory for this motherboard?! I want to fully populate it to 32GB, so really don't want to lock myself into buying 4x4GB at this time. I've found Crucial doing 'compatible' sticks @ £169.19 for a 16GB kit. Nothing at Lambdatek, Scan or Aria. Well, there's two 8GB Hypertec 'equivalent' sticks showing at Lambdatek, but I'd prefer the warm fuzzy feeling of knowing someone else'd used these in this board. Especially as I intend to go the ESXi route. Ditto for the Crucial stuff I suppose. Can anyone suggest a supplier on this side of the pond that doesn't want to charge an arm and a leg and actually has known compatible stuff in stock? Regards, overbyrn

Sorry for the stupid question, but I just couldn't find the answer in my forum searches. I recall there were reported issues with M1015's and drives going to sleep. Is this still a problem or was it squashed a long time ago? I'm still running 5.0-b12a but figure it'd be nice to move to the latest RC if at all possible. Regards, overbyrn

theone, thanks for the plugin. Been meaning for a long time to tinker with running virtualbox on unraid and this might make the process a lot more straightforward. I've just installed the plugin onto a dev box (no other plugins installed at this time). Noticed upon first entry into virtualbox setting page, there is an error regarding vms_session.cfg not found. I've only briefly skimmed your code, but I'm assuming that's because there are no running vms as yet? I think there may be a mismatch between where this file gets created vs where the php var looks for it. From the php and rc scripts: rc.virtualbox: VBOX_PLUGIN_PATH="/boot/config/plugins/virtualbox" vboxmanage list runningvms > ${VBOX_PLUGIN_PATH}/vms_session.cfg virtualbox.php $vm_session_cnt = count(file("/boot/config/plugins/vms_session.cfg", FILE_IGNORE_NEW_LINES)); May have misread the code, so apologies if this is by design but I think otherwise it may be looking in the wrong place. That being said, on first load of the gui page there isn't a vms_session.cfg in either location likely due to the reason mentioned above. Regards, overbyrn.

How can I check the validity of a downloaded file to check after download? If it is possible I will add it to the code. Calc an MD5 sum. Use /usr/bin/md5sum to compare against a known good hash.

Don't know if anyone's interested, but I've compliled a version of the ntfs-3g package against latest stable - 2013.1.13, released on January 14, 2013. Seems to be running ok on my production UnRAID box. Granted I only have one snap mounted USB drive, but so far I'm not seeing any instablilties. Side note; I was trying to track down the source of high cpu of ntfs-3g process when viewing via htop. Sorting by CPU, it constantly spikes at 100%. There are a few reasons according to the FAQ over at Tuxera and although none of them specifically say to update to version x, I thought it couldn't hurt. I couldn't locate a pre-compiled version so decided to roll my own. Thought I'd share in case anyone wanted a copy. Available here Regards, overbyrn