|
Johannes Linkels wrote: The lesson: it is not Linux which is disappointing, it is the n-th order derivative which tried to "improve" on functionality or safety.
I can't disagree with that.
My point was, I was looking forward to dedicating a machine to using Linux for a specific task. TrueNAS is very well regarded in the community. Yet it fell short enough that it put the kibosh on that project.
Obviously Linux != TrueNAS. But to me it still came across as another missed opportunity.
|
|
|
|
|
I understand. And you are right in expecting better performance. Unfortunately people are lured into using one of those specialized distros because it is "easier". Well, maybe it is for the unprepared user. I am running everything for which is a special distro on a general purpose Debian server. And true, I don't have fancy web interfaces.
The problem with reviews or comments is mostly that most products are easy to use, beautiful and perform very well as long as nothing goes wrong. The quality is in the solution is when you are able to recover from a seemingly total disaster. No one ever reviews that, if you are lucky someone posts such a recovery when it happened and if possible.
I don't want to land in a Windows-vs-Linux discussion, but Windows and Microsoft products are an order of magnitude more powerful, beautiful and easier to use. Until something goes wrong and then there are exactly zero recovery options except re-install and restore. All Linux solutions working toward a better (as in easier) experience run the risk of moving into that direction.
|
|
|
|
|
Johannes Linkels wrote: The problem with reviews or comments is mostly that most products are easy to use, beautiful and perform very well as long as nothing goes wrong.
What I'll point out here is that nothing went "wrong"; it went exactly as it's supposed to. My scenario is just not supported.
That is the part that was not brought up in any discussion I read when doing my homework. In hindsight, I would've expected that to be pretty high on the list of things TrueNAS can or cannot do.
And based on my search results (when trying to understand why only one drive was being picked up), I'm not the only one to have made that mistake, as there seems to be plenty of people finding out the same only after the fact. It's as if every discussion on the topic was just taking for granted I knew things I did not know.
Hence my remark about me "having silly expectations" (which some people seem to have taken issue with, but that's another story).
|
|
|
|
|
I like to just share my Linux experience.
Since my "system" has been steadily growing as far as shear quantity of drives due to my opinion "if it is not used why bother to delete it " one of my experiments is to play with RAID. I do realize that its purpose is to build a "backup" just in case of hardware FAILURE. With that said - I have lost access to fully functioning RAID FOUR times! Not due to hardware failure - Linux just refuses to keep my RAID functioning !
My next opinion is about how Linux, and Ubuntu specifically, handles Bluetooth... or it actually does not care about Bluetooth - in my opinion.
|
|
|
|
|
I've toyed around with RAID (both software and hardware) many times over the decades, and I finally swore off of it many years ago, when the hardware controller on my MB died.
I decided to give it another go with TrueNAS.
After this experience, I'm still sticking with the conclusion I had already drawn years ago - to do it right, RAID is expensive. RAID stands of Redundant Array of Inexpensive Disks. Yeah, the disks are cheap, I have plenty. But the infrastructure around it is expensive, if you need it to be reliable. QNAS, Drobo, Synology - I have a hard time justifying the sort of money they want as a home user.
I was under the mistaken impression that you could throw just about anything at TrueNAS, and it would just make it work. Not even close.
That was my big disapointment.
|
|
|
|
|
When one has thousands of possible hardware choices and thus the possible combinations can reach unimaginable numbers the reality is that things will not always work.
Rather surprising that there are not more failures.
|
|
|
|
|
jschell wrote: Rather surprising that there are not more failures.
How could there have been? I was blocked right from the get-go. I only wished its inherent limitations would've been made more clearly. Like I said, I thought I had done my homework and the community at large is raving about TrueNAS.
|
|
|
|
|
My experience with TrueNAS has been absolutely fantastic. I've worked with various Linux distros over the years, but TrueNAS almost hides that part of its DNA from you.
I built a NAS system several years ago, installed FreeNAS and four 3TB drives as well as a small SSD as the boot drive. Installed the software, booted it up, and was almost immediately in production. No muss, no fuss. Set the drives into a ZPool with parity, created Windows shares, started throwing files at it. Rock solid.
When the system started filling up, I started to replace the 3TB drives with 8TB versions. Took a drive off-line, physically swapped it out for an 8TB, told FreeNAS to re-silver. Array re-built, no errors. Rinse, repeat, three times. One of the beauties of FreeNAS/TrueNAs is that when the capacity of the drives in the array changes, it will automatically re-size the array after the last drive has been replaced. Again, no magical incantations needed, no chicken entrails at midnight. I upgraded from FreeNAS to TrueNAS efortlessly.
One day, the system would not boot. I don't know if it was the CPU, memory, or the motherboard, but it was completely toast. Fearing the worst, I built a new base platform, and hooked up my drives. Re-installed TrueNAS on a new SSD, loaded the last configuration backup, and bang -- all my shares, permissions, accounts, etc. were back, as if there had been no interruptions.
On a final note, I actually have two USB drives attached to the box. These are 8TB Western Digital MyBook units. They are configured as software RAID 1 (mirrored). I use them to store digital video files (no, not porn, but movies and TV series) to feed my entertainment system.
While I don't use Linux as my daily desktop, I do highly recommend TrueNAS. My main desktop is Windows, mainly because I am a Windows developer, but I do run several Linux VMs on my network for various purposes (like ISPConfig, Postfix/ClamAV/SpamAssassin).
|
|
|
|
|
Your experience is exactly the sort of thing I was hoping for.
My expectation was that it'd be fine with bunch of drives in a multi-drive enclosure connecting back to the PC over USB-C. It sounded simple enough in my mind...
|
|
|
|
|
There are a couple of things that I think need some clarification. First is that you may be mistaking Windows' willingness/ability to accept whatever drives you want to throw at it as an endorsement that what you're doing is a good idea and will perform well (both speed-wise and data integrity-wise). IMHO, that's an incorrect assumption. TrueNAS will do what you want it to do but it will not endorse it as a good idea (from a data-integrity and performance POV) because it's not. People use TrueNAS for its performance, stability, data-integrity and the UI on top of it which makes it really easy to create a reliable setup. If TrueNAS isn't letting you do something easily, that should be a sign that what you're doing isn't a good idea for a super stable, reliable and performant system. In that light, it's more of a guardrail that is intended to give you pause before hopping over it.
I think Unraid might be more of what you're looking for. One of its strengths and key selling points is that it will take whatever disks you throw at it and add them to your storage. It's also got a nice UI that makes things pretty easy to do. As long as you understand that throwing whatever kind of disks you want into your storage pool without concern for their age, quality, storage capacity, etc. is generally not going to be as reliable from a data-integrity standpoint as what you would get with better drives of matching storage capacity you'll be fine. For many use cases, that's sufficient. As long as you make sure that anything that you absolutely can't lose is backed up you should be good.
The second thing that I think needs some clarification is that TrueNAS (or FreeNAS as it used to be called) isn't Linux. It is based on FreeBSD (a Unix flavor). While both Unix and Linux support the Posix standard, they are separate operating systems with different capabilities, strengths, and weaknesses. BTW, Unraid is based on Linux (specifically the Slackware distro).
The fact that there are a ton of different Linux distros can definitely be overwhelming; it was for me when I first got started. However, I've come to view it more as giving me the ability to evaluate different things and pick the best tool for the job. I'm not stuck with taking a "jack of all trades" approach like Windows often takes. I use both Linux and Windows as daily drivers both on bare metal and VM. Both are stable and performant. It's taken me more time to read and learn about the various Linux distros but it has paid off big time in stability and being able to tailor an optimized solution for whatever computing problem I need to solve.
|
|
|
|
|
That's a great discussion and I thank you for writing down your thoughts.
Couple of things:
greyseal96 wrote: first is that you may be mistaking Windows' willingness/ability to accept whatever drives you want to throw at it as an endorsement that what you're doing is a good idea
Sorry, in case I didn't make it clear, I fully realize that Windows being more permissive and let you go ahead with it doesn't mean any of it is a good idea. I had already come to that conclusion. The distinction I was trying to make is that TrueNAS blocked me altogether. Windows makes no such attempt. But again, making that choice is left to the user (understanding risks and all).
greyseal96 wrote: generally not going to be as reliable from a data-integrity standpoint as what you would get with better drives of matching storage capacity you'll be fine
Where does that leave JBOD systems, I wonder? I didn't invent the acronym, so surely the idea has enough merit that people use such systems.
greyseal96 wrote: TrueNAS (or FreeNAS as it used to be called) isn't Linux. It is based on FreeBSD (a Unix flavor).
Someone else also brought that up, and I hadn't, because my end results would've been the same: TrueNAS Core is based on FreeBSD. TrueNAS Scale is based on Debian. I would've pointed out which I tried, if I had been under the impression the outcome would've been different.
greyseal96 wrote: The fact that there are a ton of different Linux distros can definitely be overwhelming;
I'm not terribly worried about that; one of my part-time hobbies is to hunt down random distribution ISOs and try to get them running in VMs. I'm looking at my collection right now, and the root folder stands at 607GB worth individual ISOs alone.
|
|
|
|
|
Ummm with all due respect... TrueNAS is not based on GNU/Linux. It's based on FreeBSD. Different systems and different kernels, different drivers and so on. So even if it is an open source project... it's not a GNU/Linux distro. Therefore GNU/Linux did not disappoint you yet.
|
|
|
|
|
Well if you wanna get nitpicky, TrueNAS Core is based on FreeBSD, while TrueNAS Scale is based on Debian.
I didn't bring these up because that particular point was entirely irrelevant to the discussion at hand. Are you under the impression the results I got would've been different had I been using one vs the other? If not, then again, it wouldn't have changed the discussion.
|
|
|
|
|
Debian does have older packages especially Debian Stable. FreeBSD I'm not sure because I've never used it so I assume perhaps there are differences between packages maintained by a distro vs getting the latest upstream code and just compiling it and packaging it.
Again I don't know how FreeBSD uses upstream code or what are their patching and maintaining policies so this is just an assumption and I could be wrong. If you're using TrueNAS Scale then you're correct. If you're just using TrueNAS Core then there is a possibility that my statement holds true.
Not trying to be nitpicky or difficult just making observations that might have some bearing in the discussion at hand.
I also think that perhaps you might want to look for alternative ways of making this work for you. I don't know how TrueNAS (Core Scale or any other branch of it) handles USB External storage devices but you could make a folder and subfolders and mount each of your USB Storage Devices in those subfolders. It's not as elegant or polished as the Windows approach but it's just as functional.
|
|
|
|
|
I am not sure "who is on first" AKA to whom to blame, so this is probably not Linux issue...but
here it comes...
I like to use "gparted:" but it appears to have TIMING issue with multiple partitions and anything ( disk ) large that 100 GB...
Then it keeps "scanning" , (usually) ALL devices , after minor change is made to one device....
As I am saying - it is hard to "blame" ( stupid ) behavior...
PS ...and it does not do "mount"...
|
|
|
|
|
Well, just look at the bright side. Linux prevented you from doing a silly thing. A really silly one.
Using old, used drives in a RAID, any form of RAID is simply a bad idea. It's data loss waiting to happen. Even on Windows, as you have no way to recover any data from the drives if something goes south with that array.
As far as Linux on real hardware goes, I never had a problem with video. My Linux development box is a Dell laptop with a 1600x900 display and that works just perfectly fine in Linux Mint 21.2. Some times temporarily hooking up a second, external monitor no problem either.
The only problem I really had in the last decade or so was with some Intel WiFi NIC drivers, mainly when I tried to use Fedora a few years back. But again, under Mint, no problem at all.
|
|
|
|
|
Well obviously RAID isn't a backup. I'm not too worried about losing anything on that RAID setup.
I already have 2 separate sets of backups of my main data set. All I wanted to do with this is create an extra backup set, by using drives I've retired. These drives have less than a few dozens hours on them - I used them to do my previous backups, but have outgrown them as my data set has increased in size. They've been powered on only when the actual backups were taking place, which occurred anywhere between once a week and once a month.
|
|
|
|
|
I worked as a dev on Windows for 20 years. I found it was pretty good by Windows 7 and Windows 10.
3 years ago, I switched to Linux Mint. There was a bit of a learning curve, but now, I couldn't be happier! I'm running multi-monitor dev workstation. My productivity is though the roof. Since then I installed Linux on 50 different machines with 0 issues (some servers with RAID too). I only have 1 Windows machine left, which I am about to decommission.
Then this year I got a new contract, I have to work with my customer's Surface Windows 10 computer. I'm not going to say what I REALLY think about it, but: It's full of bugs, looks like crap, multi-monitory only half work, and updates...
I was a Windows users, but now: Windows, why do you keep disappointing me?
The point is, there is a learning curve, but for most things I believe Linux has surpassed Windows now.
Christian Lavigne
|
|
|
|
|
Saw an article this morning recommending that you run the command: "Defrag C:" from time to time on your SSD drives. But does it make sense to defrag a SSD? I can understand that it is of value on old spinning disk hard drives, where fragmentation can cause the reader to physically jump from fragment to fragment, but a SSD has no moving parts.
What do the experts say?
Ok, I have had my coffee, so you can all come out now!
modified 3-Dec-23 8:47am.
|
|
|
|
|
No - all it will do is 'burning' write cycles, which shortens the life of the drive
"If builders built buildings the way programmers wrote programs, then the first woodpecker that came along would destroy civilization." ― Gerald Weinberg
|
|
|
|
|
This is a well-known and probably correct argument.
On the other hand, the question I have is: If SSDs use something like DMA, could a certain kind of defragmentation increase throughput?
|
|
|
|
|
Agreed.
Steve Gibson (author of Spin-Rite) has discussed this numerous times on his Security Now podcast, and it makes zero sense to "defrag" an SSD.
Some people have called him a quack, and I originally sided with them (somewhat), but after listening to his podcast for nearly a decade, it's clear he's technical to an extreme and very knowledgeable. When he does a deep dive into some technical matter, I think he always makes a lot of sense. He's not clickbait-y and doesn't make outrageous claims.
Not that I had any doubt, when it comes to defragging an SSD. But his explanation for it (I don't have a show number for it, sorry) just sealed the deal for me.
modified 3-Dec-23 10:35am.
|
|
|
|
|
dandy72 wrote: Some people have called him a quack
dandy72 wrote: it's clear he's technical to an extreme and very knowledgeable
You can be both you know. 
|
|
|
|
|
Fair point. But I've been listening to his podcast for over a decade, and I have come to the conclusion that those who called him a quack were just poorly informed.
I forget what his exact concern was (something about XP's default network configuration?), but in the end he was proven right and Microsoft eventually had to seriously lock it down with SP2, which introduced (for the first time) the Windows firewall.
|
|
|
|
|
I remember him alright.
He isn't a quack, but has a tendency to fight windmills. 
|
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
