SSD

Sorry, realizing this was the binhex thread. I am running the LinuxServer version. Is there a similar link for those tags?

I've had issues with 2 Medusa updates, and am generally nervous about updating. Is there a simple way to apply an update and test it out, but then be able to revert back to the prior version if there are issues?

I noticed this happened on a new build I am doing. It applied to system, appdata, domains, and one other I think. I deleted them from the array disk. But I think this is probably something @limetech would want to hear about if it is affecting others as well. I actually wound up losing a VM XML in the shuffle, but it was easily recreated as I had the image file. But other users might not know how to do that.

Passthrough is a VM concept. Never need to pass through anything to a Docker. What you need to do is set up a volume mapping of the UD to the Docker to be able to see it in the Docker.

Nice workbench concept. Well executed. But ... If its like my basement, there is storage stuff coming and going, dog winds up down there from time to time, xmas decorations moving in and out once a year, looking for tools, rummaging through old boxes of junk. This is too open and easily knocked by an unsteady human with awkward load. One steadying hand lands on one of those drives and it is going down, getting twisted and mangled due to the screws trying to hold it vertical against an adult body weight. And then crashing into its neighbor with metal grinding against metal and a spark or two as the electrical cords get pulled out violently and cables rip from their connectors. Mangled fingers, cuts, screams of pain, emergency room, fire department, law suit, wife on the warpath (understatement). Massive data loss. Who knows! And if it is you that did it, no one else to blame! I can see no happy endings with one accident. Maybe if the top would flip over and drives dangle down. That would help. I'd like giant cover for this thing with fans blowing in and out that could be removed when you wanted to work on it. Would block balls, pets, kids and other humans - generally make it a lot safer. But leaving it like this 24x7 is a liability concern and puts the data at risk. No way!

I think this newest method is going to take care of the problem, but would not want to our brain power coming up with ways spammers could circumvent. Let's keep those to ourselves and not share with the enemy. Feel free to PM me if you have specific concerns or ideas along those lines to shared among the moderator / Limetech community,.

@pwm The mods are experts and have been reading the forums for years. It will be very difficult to fool us with old or made up posts.

Just marked 6 spammers in approval queue! (No legit new users to approve) It's working!

I approved one new user post. There was no spam in the approval queue.

Ahhh, makes sense. How did you mount - did you use those somewhat flimsy metal strips I mentioned?

Ahhhh. Makes sense. I'll look in my spare parts I probably have something similar. Don't know if I'll use it for spinners, as I want the hot-swap for them, but sounds good for SSDs. But I do have 2 RAID0 drives for parity. Might consider putting them there if I can get good cooling to them. They almost never need to come out or be swapped around. And cabling just two drives not much risk. Yea - too bad they didn't think to put a 120mm fan there. Disadvantage of a case from that era, when 120mm were considered large, and 80mm more the standard. I have a 4.5" hole saw. Planning to drill the fan mounting holes in the top of the case, and use them to screw a piece of plywood to the top of the case. Then use the plywood to drill in the arboring hole. Should work and keep the hole saw from bouncing around as it hits the case cutout for the 80mm fan. Once cut I can unscrew the plywood and mount the fan. While writing this I thought about moving to a 140mm fan on top. Have to check to see if there is room. Can you clarify? I think there is only room for one 200mm fan. They are huge. Are you using 2 120mm fans mounted in that round cutout? Think it might fit, but not sure how you'd secure them. The 200mm can screw to the rather flimsy metal tabs that are present, but not sure if it would be good enough to not rattle. Probably would need to secure it better. On my setup, my 5in3s each have a strong 92mm fan. So I'd have 4 of those + 200mm (100 cfm) fan blowing in, and two 120mm blowing out, (one back and one top), maybe 80mm fan in the lower PSU area, as well as the PSU exhaust. I have some rather loud 120mm fans that move a lot of air, but even using them there is still going to be a lot more air coming in than going out (not optimal). If I omit the 200mm fan, the in and out flow would balance better. I already have the 200mm fan, so may try both ways and see what happens to the temps under load. Still considering if there is another location for another exhaust fan. So you are using the 80mm hole for something. See my post above. Thinking of just have two pushbuttons in the black plastic. Or use the intrusion switch rigged for as power initially. Thanks Johnnie!

Yea - that's what I was thinking. Found these, that are a little pricey but would mount nicely on plastic panel at top front. https://mnpctech.com/case-mods-gaming-pc-liquid-modding-custom-computer-mnpctech-overclock-cooling-fan-grills/switches-toggle-pc-computer-power-reset/red-vandal-resistant-pc-power-switch-22mm-clone.html Other idea was to disable the case intrusion switch, and use that for power. Don't think I much use the reset button on the unRAID server. Used it from time to time on old Windows boxes. This would be super secret. If off, no one could figure out how to turn on!

Few questions if you wouldn't mind ... 1. Can you explain the mod you did? Hot swap or fixed installs? Cooling fan? Have a picture by chance? Too bad the opening is so short. 80mm fans are pretty small. 2. I am removing the 80mm top fan and replacing with a 120mm exhaust. Took a little dremeling on the upper PSU cage to remove a cross member to make room. Expanding that 80mm square into a 120mm (4.5" actually) hole is not going to be easy, but have a plan. Did you do a similar mod or leave this fan alone? 3. Considering adding 200mm fan to the side panel for extra cooling. Did you put a fan there? (If so, input or exhaust). Read side fans in general may not be helpful. But with 2 GPUs + more powerful CPU in there, might need extra cooling. 4. Getting the 12th (top) front drive slot free required removing the piece that includes the power button, reset button, lights, and usb slots. I don't care about the usb, or even the lights, but definitely need power and reset. What did you do? Thanks man!

I found one of THESE brand new on the clearance rack of a mom and pop computer store. $75 (have a $170 price tag on it). 12 drive slots (enough for 4 5in3s (20 drives)). Lots of aluminum. On wheels. Made like a tank! It is taking a little modification with a dremel and hole saw (to enlarge the top fan), but I love it! Looks like the Antec is going to be retired soon!

Some motherboards look for a drive to slap an HPA onto. You might have ongoing issues. Suggest disabling. But if ever the battery goes or you reset the settings, the old default setting comes back, and you can have troubles again. I believe gigabyte changed the default to be disabled.

@bombz - You will need to do a new config to remove the old drives and add the new one. Don't delete the data from the 2 500G drives, and you'll have a backup just in case you have any issues with the external. I do recommend a preclear to test out any new drive, but technically you would not need to preclear it. When you did a new config, parity is rebuilt. What you could do is this. It's what I would do. You don't need the external. Preclear the new 4T drive and second parity disk (if you have). (Recommended but not mandatory) Shutdown server. Remove one of your 2 500G drives from the server, and add your 4T drive. Boot server. Do a new config but keep configuration. Add the 4T add as a data drive and omit the 2 500G drives from the array configuration. (I'd also omit parity, but that's up to you. Without parity you'll be able to copy faster). Change filesystem to xfs on the new 4T drive before starting array. Start the array. (Let parity build if you added parity.) Format the 4T drive. Use unassigned devices and mount the 500G drive. Copy all the data from the 500G drive to the 4T. After the copy, shutdown the server. Swap the 500G with the other one. Boot the server, start a array, use UD to mount the other 500G, and copy other 500G drive to the 4T. Shutdown the server again. Remove the second 500G. I think you said you wanted to add second parity drive. This would be the time. Otherwise just leave slot empty. Boot the server. Before starting the array, add your parity disk to the config (if didn't do before), and your second parity if you are adding it now. Start the array and parity(ies) will build if you added.

@bombz - So it sounds like you are completely full in terms of drive slots. So you want to copy data from 2 500G drives to an external drives, and then add a 4T disk that you'll format as xfs. This sounds like a reasonable approach. It will require rebuilding parity to remove the 500G disks and incorporate the new 4T disk containing data. Swapping add-on controllers is seamless. UnRaid recognizes drives by serial number. So if you remove the old, attach the drives to the new, and boot, all the drives will be assigned to their proper slots as though nothing had changed. (Obviously connecting everything securely and not disturbing the drive cabling of other drives is critical to not causing drives to drop offline). Related to ports to avoid I am confused. The SuperMicro add on controllers that use the Marvell chipset are problematic, but motherboard ports that are provided from the Intel chipset or non-Marvell chipsets are fine to use. Your post made it sound like you were trying to avoid using regular motherboard ports but maybe I misunderstood. I do suggest planning for an empty slot as it is useful in some recovery scenarios. And comes in handy for preclears as well. It is also needed to convert from rfs to xfs.

The "U" is tilted and gives me impression of lack of precision. I draws the eye, but maybe in the wrong way. But I do like the linked U and N looking like a chain. The stacked drives on the "I" don't really come across as anything disk oriented. The D looks funky. THe "un" sort of gets lost. Heavy use of stripes - when unRAID does not stripe data. Maybe not a good fit. I do like tagging with Virtualization. With VM pulling as much or more attention than NAS function, a tagline that pulls it all together might be helpful. "The Ultimate NAS/VM/Docker Platform."

The expander can be mounted anywhere, inside or even outside the case. Just needs power and SAS feeds from HBA - and to not short out. It does not need to plug into a slot, although that is a convenient place as it supplies power and avoids need to mount it somehow.

Thanks @dmacias!! I didn't remember I had loaded Python through nerdtools - thought it was stock. Now I remember where it came from! But I just installed "pip" and it installed the pyquery package like a champ! Too late to play with it tonight, but will have a go soon and report back. Thanks again!!

Other thing to consider is space to do any type of file recovery operation. The recommends minimum free space setting is 2x the size of the largest file on the disk. That is room for one more file to copy + room to recover a file. Like I said, this is bare minimum.

I am on unRaid 6.4rc6, and the python version is 2.7. Not sure what version comes stock on later RCs. I've had no reason so far to try to move to Python 3. THANKS!

@dmacias Hope this is not out of bounds for the devpack team, but I have been playing with Python and would like to install a python package. I found it referenced in a pretty cool script that will create an IMDB category in a Plex DB for the top 250 movies of all time (any that you own). It scrapes an IMDB web page and parses it with a Python package called pyquery. I would like to run that Python program and have some other uses as well. Here is a link to the script https://gist.github.com/shariat/313795d59931c4d128d2 You'll notice it contains this line: from pyquery import PyQuery as pq And here is a link to info on pyQuery: https://pythonhosted.org/pyquery/# And here is a link to SlackWare packages for Pyquery. (Notice it has dependencies that have dependencies ... probably need 5 or 6 packages to satisfy all those.) https://slackbuilds.org/repository/14.2/python/pyquery/ Anyway - I am a bit clueless how to get this installed in unRAID. I DLed the packages and tried to figure out where to put them and how to make them do their thing. It all looked so easy! I unpacked them, tried pkginstall, and coping things here and there. But I didn't have the special sauce to make them dance! They just sort of stared and me and laughed . Any help appreciated!! And if it could be added to the DevPack, that would be awesome! Thanks!!

Assuming you installed the Docker and re-established your libraries. Virtually everything in Plex is run from a database which assigns ID numbers to each item. These IDs will be totally different between your old database and your new database. Even the library ID will likely be different. It would be much easier to find your playlist settings and just re-establish them in the Docker.

@johnnie.black - If a file is corrupted, it seems you are totally blocked from accessing that file (for partial recovery). Is that true? To others considering what filesystem to use and like this feature, some people run tools that capture md5 or similar checksums with XFS. It's not real time checking, which has advantages and disadvantages. I run something home grown, but there are tools in CA that you can install. When I decided on xfs I was a little concerned that btrfs was not able to report free space accurately. The fact that it was more easily corrupted with a hard shutdown, and did not have nearly the user base of xfs were also factors. Also never realized why it needed ECC memory morso than any other filesystem. All this made me feel it was more fragile than xfs. I'll try beta this and beta that, but for my filesystem I definitely am very very conservative. I was bitten by a file system bug in RFS, that was losing users in droves, and, going forward, I want rock solid with huge install base including production systems. Those guys will make darn sure that the filesystem updates are extremely well tested. I continue to monitor what Johnnie has to say about btrfs, and positive experiences of other users. Have not read horror stories like I was worried would emerge. But I'm sticking with xfs, for now anyway. But I do kick the btrfs tires from time to time.