Ubuntu users FYI: You can freeze any package including kernel packages by: example-1: sudo apt-mark hold example-2: sudo apt-mark hold linux-image-generic linux-headers-generic
Quick tip: When you type a command that should have been done with 'sudo' but you forgot it, you can type: sudo !! This will rerun the last command given under sudo without having to retype the entire command.
with Bleachbit never run free hard drive space on an SSD! You'll unnecessarily write zeros to your drive, thus wearing it out faster. On the other hand it's extremely handy to do that now and then on mechanical drives.
Ah, you forgot to add that on Arch-based system we also need to: - clean pacman cache (all downloaded packages are stored there and taking space) - uninstalled orphaned packages or packages that were removed from the repo (and we cannot do that blindly, because it often happens that packages are marked as orphaned mistakingly) - update our configs with pacnew files, this is a manual intervention that is needed to be done once in a while - update git packages if you have any - they rarely show as an update and this is usually a manual action, although there are commands that can mass update git packages
One correction: 'apt-get autoremove' removes packages that are no longer needed i.e. packages that were installed as dependencies for another packages that were removed and these packages were left of and are no longer needed. On Arch there is no such command. You'd have to combine 2 commands: 'pacman -Qdtq' which gives you the packages that are no longer needed and remove them with 'pacman -Rsn' or combine them into 1 command: 'pacman -Rsn $(pacman -Qdtq)'. Also if you use Arch it's probably a good idea to remove packages with 'pacman -Rsn' which removes packages that were installed as dependencies for the package you remove and are not needed by another package and also remove their config files in /etc, instead of 'pacman -R'.
I would say, be careful with what you remove with pacman -Rsn as it is entirely possible to accidentally get cross-dependencies and end up removing something that you actually NEED on your system. Source: I've done this by installing a Gnome tool, not thinking about it before issuing the command, and wiped out my entire DE including my network manager, so I had to fix it from a chroot.
If you use the APT package manager, I would recommend using nala to upgrade your packages. It's meant to be more user friendly, has a cleaner terminal output, and automatically does certain things for you (autoremoves/autopurges files that are dependencies no longer needed, automatically updates the source list before upgrading so you only need to run sudo nala upgrade, etc.). You also aren't replacing apt, so you can still go in and run maintenance scripts without needing to rewrite them first. Of course, the apt command family is going to be your de-facto option for more advanced operations, but with nala upgrading is simple and enjoyable.
Chris Titus Tech: Chris. Just a tip. I was always told to never stop bleachbit when it is in the process of cleaning. That you must let it finish or you will have problems.
That's really neat. I didn't know that you can keep Linux running indefinitely without a reboot. It's kinda ingrained into me that a computer is going to need a reboot eventually, because of memory leaking or whatever. That's mighty impressive!
Yeah I hear you! I still reboot often because I change tons of things, but yeah... It is not unheard of that many closed systems like elevators, electronic posters, etc... have run linux on a battery backup for 10+ years.
Yea, I know it can be run indefinitely but I still shut down my computer every evening to control my electricity costs, just like I turn off all lights when I go to sleep.
I’ve had macOS with an uptime of nearly three months as well. Memory leaks can happen in software, but especially in the OS it should only be in corner cases. If it’s not in the OS, the OS should be able to clean it up when the offending app is quit.
@@ChrisTitusTech I've just started using Ubuntu and I'm new to Linux. Love your channel. I've noticed that every time I reboot, especially shutdown, I get the black screen and there's a lot of broken code running across the screen. Then it shuts down or reboots. Goes by really fast and bothers the crap out of me. Is this an issue with the bootloader? Any ideas?
I love conky. It's one of my favorite add on apps. I love configuring it lol. It can be an insanely useful tool for monitoring your system...and it looks cool.
I find it easier to manage the repositories from synaptic (for a debian system). Just pull down the menu and tick/untick the repositories you have, you can also add new repositories and delete old ones. If you want to remove a package it will show you what else you would need to remove at the same time. There's really no need to edit /etc/apt/sources.list directly. I also almost never reboot after an upgrade, if there are new graphics drivers, I tend to just kill Xorg and that does the job.
I don't recommend doing most of these except verifying your backups, installing updates, and removing no longer needed and orphaned packages (dependencies of older version of packages that are now no longer needed). Especially that thing you do with Bleachbit or similar tool is going to cause more problems than it solves. Unless you have a particular problem with a specific application, you generally shouldn't micromanage these cache files. It won't make your system any faster, and more likely than not it will just slow down programs and cause data corruptions. In low disk system, it may be worthwhile to clean the package managers' download caches, but if you aren't running low on disk space, you don't need to worry about it. One maintenance that I recommend that hasn't been mentioned, if you're using Ubuntu-based system on btrfs system, you may need to remove the automatically created btrfs snapshot that apt created before a major system upgrade. Once you've confirmed the upgrade is successful and you decided you likely won't be rolling back, you can just delete the snapshot using btrfs tools. apt won't remove them automatically, and these snapshots take up quite a lot of disk space (and because of copy on write, they'll take even more space the longer you keep them).
Yeah once in a while I'll just see how much storage space I have left. Linux can get loopy if you get up to 90% full. Up to then no worries though. Right now my root partition is at 32% so no problem there. If I get up around 80% I'll think about maybe ditching something then. Or I'll just get more storage space. Probably the latter knowing me.
Greetings, I have been transitioning to Linux exclusively for a few months now. This video has been very helpful. This and your other videos have been very helpful. Ciao!
You do the best tutorials Chris! Clear instructions, reasons why you're doing stuff, great explaintions and poiniant tips. But best of all is your smooth delivery and non-intimdatng predisposition. Love your work mate, it gets me pumped about the future of the world. Thank you kind sir. 👍💯
Many of my users are running my software on Win 2000 still. No maintenance required, but using very few Windows facilities. POS/Inventory and such. Linux is better, but less familiar and has fewer available third party programs. In Windows, I wrote my own scheduler as I did not trust the one that came with it. This can be done. There is push back on Linux desktops. Linux servers are well received, is my experience.
On the linuxmint forums, they warn that bleachbit is bad, that it can break your system, and that you shouldn't use it. I have used it for a while and I have had no problems.
Thanks Chris for warning users about doing a backup before purging packages, purging packages is a very dangerous operation, you can mess up your system beyond recognition.
5 лет назад+7
6:18 and 12:43 when you forget "sudo" in front of a command, you can look 110% more cool if you run "sudo !!". It's a proven scientific fact.
5 лет назад+1
@callam85 Ladies like men who use "sudo !!". You're just jealous.
Just to expand on the no need "no need to defragment" part on linux. The reason linux and most unix/bsd filesystems do "not need" defragmenting is because they leave more space between files to allow them to grow on the block level unlike ntfs which uses the immediate space (pretty much) after a file is done being written so any newer data for the file has to be written elsewere and a reference pointer is added to the end of the old data ("continues at XYZ") rinse and repeat, however you can run into situations where defragmenting the filesystem will be needed, however it is done in the background and a manual run might only be needed when the disk is close to being almost full (which in most cases will be rare) so yes you dont need to defragment your system like you have to do on windows unless your disk is pretty full before you will need to do a manual run (we talking maybe 80-90% full) before that, all defragmenting is done as needed in the background without you noticing any impact from it but you have to go over 50% disk space used by average before defragmenting is even done in many cases, as no files will be fragmented (for the most part) before that limit is reached.
HikariKnight keep in mind that SSDs should never, ever be defragged. Not only does it waste lifespan, but if the controller is cheap you can brick the drive.
@@jadosrd8950 If this ["brick the drive"] would be true, any IO access pattern could brick your drive: You should just not (and don't want to) use such drives. There are applications where "defragging" an SSD can make sense but that needs special defraggers that care about reducing IOPS for your specialized application and not moving files around as most classical defraggers do (moving files around is pointless on SSD, and introduces additional write load). But "never ever" is also: Your SSD wouldn't just die the next day just because you defragged it yesterday. It lost maybe 0.5% of it's life span, on modern drives probably even less. The bigger the drive, the less prone it is to this problem: Bigger SSDs have a exponentially longer live span for non-suitable write patterns.
Kai Krakow again back to the controller chip. A good, or even half decent, chip can withstand defragging, but it also makes it a moot point as THAT IS it’s JOB to read/write a bunch of scattered data as if it were in one chunk. A bad controller, and Im mainly talking $10 Ali junk here, has been shown multiple times to not “keep up” with knowing where data is and subsequently “losing” important data (Windows files).
Makes me wonder why Microsoft just doesn't create their own distro like RedHat. I mean the biggest thing is software, but if they created it alongside windows, then use something like a .NET virtualized container, while supporting developers that create main ports to their vision of a linux OS. I mean powershell and windows server is a huge hurdle to overcome, if they were to do that, because powershell is object oriented and linux is textual. Windows Server has implementations all over enterprise environments. So actually fundamentally changing Windows may be too volatile for them.
"FAT is event worse...". MS has just open sourced the exFAX spec and is supporting exFAT being added to the Linux Kernel. Let's see if MS will pull this off. I think they will.
Awesome video man, this is the kind of stuff I'd love to refer to people who are just getting started with Linux. I would also add `htop` to the list too as for me it stands in place of windows task manager. Love the channel. :)
I'm currently doing a personal experiment with my Manjaro KDE system. After a couple of update breaking (that i restored with Timeshift), i decided to completly stop doing any updates whatsoever. That was back in May of this year. So far everything i have installed back then work fine. I did install one or two other programs that work fine, but some others i could not because the dependancies are not up to date. I have another system on another SSD, Pop Os, that is up to date and can utilize for whatever reason. But in conclusion, my Manjaro still works like a charm and i use it every day as my main system!
In addition to `apt autoremove` I do recommend running `apt-get auto-clean`. It looks like `apt` likes to hoard every downloaded package and that could easily get over gigabytes. I had switched to 'testing' that had regular updates. It has gotten around 8GB of cache... on my 32GB sdcard. To empty the whole cache you can do `apt-get clean`, this however would remove the archives of the currently installed packages too.
Oh come on, Chris... I'm a fan of Linux, too. And I don't like Windows for a lot of the same reasons as you. But saying, that NTFS is just a slightly improved FAT file system is just technically wrong. In fact, NTFS is an extent-based FS just like ext4, and it has a journal similar to ext4, maybe better compared to XFS. And what's also not true is that defragmentation is needed in NTFS and not in "Linux". In fact, neither needs defragmentation, both do - thanks to extents - a pretty good job of keeping fragmentation low, as long as you don't fill your drive too much or have some very strange file access patterns. Classical defragmenters are mostly useless. What should be done, tho, is using an advanced optimizer which puts your files back into a better location (i.e. "MyDefrag" can do this, and even the Windows-integrated defragger does this in some manner). Why is it needed? Because Windows Update messes with the optimal placement of your DLLs a lot, this slows booting and application starts a lot over time. This is not a fragmentation issue, it's a file placement issue. And guess what: Linux suffers the same problem, constant updates mess a lot with file placement over time. This is why there exist boot-time read-ahead programs, and boot-time file location optimizers (i.e., I think e4defrag can do this). BTW: File location optimization is not needed on SSDs, just do not care about this when you're on SSD. But still, Windows can suck a lot more when it comes to hard disk IO, and this is mostly not due to fragmentation. This is because Linux has quite a good IO scheduler, while Windows mostly doesn't (it's a very simple one and doesn't care much about fairness, it just looks at which processes are interactive and which are not, I think). In fact, MS thinks this is not a problem the OS should solve because you can either use SSDs or advanced RAID controllers that do it for you. But again, this is not a FAT vs. NTFS vs. Linux file systems problem, it's an issue baked (or not-so-baked) into the OS itself.
Yea, the IO scheduler in MS Windows is in desperate need of an overhaul. Since anti-viruses just seem to take top priority... which just seems to kill performance...
THANK YOU. NTFS has been around a long ass time and it's been far and away better than FAT32 for a long ass time. This seems to be a common theme among Linux users. They think the FS Windows uses is this slow outdated file system when it's actually incredibly competent and has been steadily updated for years now.
@@arnox4554 The drivers available for Linux to mount NTFS partitions are terrible and combined with the IO issue in Windows. It creates the perception to Linux users that overall, NTFS is terrible. The NTFS drivers are terrible as they have to essentially be created with reverse engineering and little to no external or, it seems, internal documentation is available to the public or appears to exist at all on how exactly the NTFS works under the hood.
@@hanro50 Yeah, that's what I was suspecting too. I think ext4 also suffers though under Windows? I'm not sure, but it really looks like if it's not native ext4 or native NTFS, the performance goes down the toilet.
@@arnox4554 I think it comes down to a simple fact: The IO layer of Windows is designed much around the design of NTFS, the single file system of Windows (in terms of "as the system drive"). That may have some advantages compared to Linux which has a lot of filesystems. So it's probably expected that any non-native filesystem will perform exceptionally bad in a lot of corner cases. But that may also be true for native filesystems, i.e. I'm using btrfs and I like it a lot for different reasons. It's not exceptionally fast because it's copy-on-write. This is not my issue, you can observe the same problem with ReFS in Windows. But let's look at btrfs integration: It's not a traditional block-based file system as, e.g., ext4, it behaves more like NFS (yes, not NTFS, I mean the network file system). Why is that? Well, there's not a direct mapping from the block device to the FS, that makes it a virtual file system with virtual block devices: If you mount a btrfs, the device file acts more like a network address than a block device because btrfs supports sub-volumes and multiple devices. That means, as for any virtual file system (aka not mapped to a block device) that we have unstable st_dev and st_inode members in the stat syscall: Across reboots, the returned values may change. It's only stable for the duration of the mount. This can confuse a lot of software, i.e. baloo in KDE or backup programs that detect differences by values returned from stat syscall. What could we say? The virtual filesystem layer in Linux wasn't designed around the idea of having a non-1:1 mapping between block devices and filesystem. That's the kind of design I'm talking about. I'm pretty sure the IO scheduler of Windows has a lot of assumptions and design choices build around NTFS implementation details. That allows it to be much simpler while at the same time being not too bad. But there's room for improvements on both sides. So it's not a question of which FS is better, it's a question of integration and design. NTFS was made for Windows, it's not a POSIX filesystem. Ext4 was not made for Windows, it's made for the Linux VFS design. I found NTFS to be incredibly stable in Windows (given the right underlying storage layer), be it performance or reliability. I cannot actually say the same for ext4 in Linux (tho, my experience is some years old for this), I found XFS to perform way better. Also, NTFS has gained a lot of features that Linux filesystems are still missing: Online hot spot repair (repaired while running without requiring a full chkdsk scan), offline hot spot repair (for what can't be fixed online but also without full scan so it's fast and silent), volume snapshots... Btrfs tries to do something similar but approaches the problem from a different direction (similar to ZFS). ReFS has integrated multi-tier support (you could achieve similar results with bcache + btrfs, and bcachefs is on the way to integrate something similar). There's a lot of exciting stuff going on currently, and in this regard, ext4 is just from the stone-ages (but also "rock solid", haha, stone, rock, you get it?), it hasn't gained the same amount of feature improvements than NTFS although they seem to be based off the same ideas and were designed at around the same time, I think. So, let's conclude: What is good and what is crap is mainly a question of expectations. Even fat32 can be totally perfect for your application. But almost always something may become crappy if used in the wrong way. From this perspective: ext4 is quite crappy for most of my applications.
Two things I noted here. 1. As a general rule don't pin packages on Arch. Partial upgrades are unsupported and are prone to causing system breakage. 2. When removing packages on Arch use sudo pacman -Rns instead of just -R option. That will also remove any "orphaned" dependency packages and leave you system much cleaner. Edit: Rephrased for clarity.
I love your videos, but as a brand new user, it would help me if you slowed down just a little bit and included a few explanations of some of your terms. Otherwise, you have been a lifesaver for me!
Bleachbit is too aggressive and too complex for an average user. I thought I know what I was doing but after some time I discovered some small breakages, lack of data that most probably was "cleaned" by bleachbit and I was very, very careful and set only most general and basic cleanup and yet it was still too invasive.
I am not very fond of command-line. I am not afraid of it, I just use GUIs whenever available. Call me lazy if you like. I used Mint Software Manager to install Bleachbit. So, let's continue with the video.
Chris you should never, ever hold or freeze packages or do partial updates in Arch Linux, madness lies that way, its like feeding a Mogwai after midnight, just don't. one update tip I have is to install reflector and set up a service ### [Unit] Description=Pacman mirrorlist update [Service] Type=oneshot ExecStart=/usr/bin/reflector --age 24 --protocol https --latest 10 --sort rate --save /etc/pacman.d/mirrorlist ### Then set up a bash function in ~/.bashrc ### function UD { sudo systemctl start reflector sudo pacman -Syu #yay } ### then save and source ~/.bashrc then just type UD in konsole to get the most recent updates from the fastest source repo ....or if a fellow KDE user install discover, appstream-qt & archlinux-appstream-data (and deps) so KDE notifies you of and can apply updates for you.
@@PenguinRevolution not really, one of the main provisos of using Arch is no partial updates/ held packages, do so and at some point you will regret it. Arch rolls and packages can move from core to extra frrom extra to community to AUR and from AUR to orphan, and vice versa, without warning or notice, if you have frozen a package and that happens to a dependent or dependency then.... Check out the arch wiki pacman entry and search for 'partial', while pacman.conf has the 'ignorepkg' and 'ignoregroup' sections its probably not wise to use them unless you are using a private repo you control the packages within and know they have no dependencies outside of that private repo.
@@mrmysteryguest I understand the concept of rolling release, so you don't need to educate me on that. I stand by what I said, it depends on what you are holding. You can get a way with a lot if you know what you are doing. You can hold your kernel without any consequences, there was a time when that was a requirement because updating kernels could break your system. also if you holding something like GIMP or LibreOffice then you probably can get away with it also. If something breaks all you normally have to do is un-hold your package and update it. I've used Arch in the past and I've held packages back and never had a problem.It's also potential problems like this is why I say running arch is never a good idea anyway, at least not for most people.
@@PenguinRevolution if with the last third of your post you are saying, "I used to run Arch, I held some packages, updated, had problems then stopped using it and advise others its not stable and breaks a lot', I refer you to my original post about feeding Mogwai's Also historically holding GIMP was not a good idea due to its connection to the GTK Libs If there is an upstream issue with a package like a kernel, the idea is to carry out any instructions on the Arch news page - eg to use the live media to chroot into a broken install to roll back using the pacman cache while its fixed upstream (extreme example), I've known this to happen but this is not the same as holding or pinning a package in Arch, which always brings issues in the longer run. Most ex users I know who had issues with rolling distros did so because they stepped outside of the the distro's recommended processes often by carrying out partial updates or pinning packages (thats not just Arch)
@@mrmysteryguest Arch is unstable in nature. everybody has breakage at sometime. THat's why it's never a good idea to run arch as your daily driver. In my opinion Arch is an unstable mess.
Ironically one of the things I learned from this video is the package name for kcharselect. I really badly needed a program like this and Kubuntu didn't ship with it. Also tried bleachbit and got 20GB of space back, which is great :3
im relativity new to the linux world , but i believe my pc experience carries through .. for clarification i remember when Microsoft released there first operating system and have used cpu's measured in hertz not kilohertz or even gigahertz! and ram was measured in k's so i believe that dates my usage experience. from my experience all os;s degrade with user input.. all os;s have recommended maintenance.. all will sorta fix the issues and take time... installation of every os ive ever done takes around an hour solving cleaning and troubleshooting always takes longer,, having an archive of all drivers and software installed says if data is saved to non stand locations a full install of the os and software is less effort and time than "maintenance" ever was and has far better results for me anyways mileage may vary
Chris I believe Fedora uses DNF package manager instead of YUM, REL/CentOS still uses YUM for the moment but with REL 8 I believe they will ship it with DNF. Also to complete the update you need to run sudo apt upgrade.
@@JohnMiller-mmuldoor A lot of people get that mixed up these days (I do it sometimes too) The only reason REL doesn't use DNF is because Red hat hasn't had a major feature update since DNF was released. and REL and CentOS also provide you with a path to replace YUM with DNF if you so choose, however most people don't and in my opinion if you do use REL or CentOS then it's not worth switching it out until the next major release comes out, YUM works just fine.
/tmp is not cleaned by the system. Linux does not delete your files unless specifically told to. The impression the /tmp gets cleaned upon reboot comes from the fact that on most distribution a special filesystem called tmpfs is being mounted at that location. tmpfs is a special filesystem that lives in RAM rather than on disk - that's why it gets cleaned when the system shuts down. Linux has nothing to do with it. If you tell it to mount a disk partition at /tmp or not to mount anything there at all, it won't and your temp files will then survive the reboot.
Had to clean the cache every couple months or the root partition would fill up and I couldn't update my Arch. Started to annoy me, so decided to mess around in package manager and keep less copies/versions. Now it''s fine.
For Fedora 31 you will want to run the following in the terminal prior to launching bleachbit once installed: sudo dnf install python2-gobject -y It works fine once you do this. This is from their News segment on their site (link: www.bleachbit.org/news/bleachbit-30). Enjoy!
I don't know anything about btrfs, zfs and HAMMER, but it appears to me that there is some performance increase after defragmentation a ext4 volume. Ubuntu even comes with e4defrag utility preinstalled. Here is the article I am referring to: www.linuxsecrets.com/kdocs/ols/2007/ols2007v2-pages-179-186.pdf
Many People duel boot so they will be using NTFS. Timeshift is not compatible with Linux Ubuntu MATE, at least not on my PC because it screwed up my system and broke it when I did a restore.
Some Linux hardware h has to be defragged. SSDs and M.2 drives need periodic trimming. This is usually scheduled by Linux on a regular basis such as "every 200 hrs of use, or after xxx gigs have been written. Same requirement for Windows systems. Eventually, the SSD bios will have an extension to where the SSD itself, without Linux, will do its own trim.
I use Discover and Synaptic Package Manager to uninstall broken PPA and residual config. Also with the autoclean, autoremove commands. However I'm reluctant to use Bleachbit, had problem with it in the past.
I guess I'm just lazy because I don't do any of this. I just run aptitude update;aptitude safe-upgrade occasionally. As long as I'm not running out of disk space I never bother uninstalling anything. I'll make sure nonsense processes aren't running. But I could care less what's wasting disk space. I got plenty of that. So maybe you can do a video about how you examine running processes? pstree is cool.
NTFS is not crappy, pretty modern FS with few drawbacks. With proper usage, it's perfect for specific tasks. I have done a ton of research on file systems, and ext4 also has lots of weaknesses, for example inode allocation, no cow functionality etc. Every major Linux FS has some kind of defrag... its just how rotational media works, after some degree not much you can do to avoid it. I keep on my eye on SGI's XFS and Btrfs, in the future they have the potential to be the next big hit under Linux.
Krisztián Kovács ntfs is nowhere near modern (1993). It was designed for low latency microwrites on striped arrays in NT Server 3. It just happened to be more modern and more stable than FAT. Not saying it’s “bad,” but the only reason it made it to consumer Windows was lack of a better option for large drive support
@@jadosrd8950 I'am talking about NTFS 3.1 (NTFS 5.1) It's not fair to compare ext4 with ntfs 1.0 from 1993. But if you compare NTFS 1.0 with Ext or Ext2 from the era it's not that terrible. In the SSD age no one cares a about fragmentation, and as is said before ext2 was also terrible. In 1993 many ways inferior compared to NTFS.
Linux systems can get fragmented if you have very little disk space, but otherwise it almost never happens in ext4 (which is the defacto fs on most distros)
Well it still happens with a lot of disk space actually. It’s just that it doesn’t happen for long, because when the disk isn’t super active it’ll reallocate things (pretty quickly) to fix it itself
Hey Chris. Just noticed that this video is from 2019. Can you make an updated version of this? Like maybe there's something new or noteworthy that you can show us in 2020. Thanks!
should do a: when i upgrade to a new major change distro version. Something i noticed when upgrading from pop os 19 to 20 it broke alot of app installs. i use alot of different tools to install: pop shop, apt, flatpak, snap, chocolatey. it can really mess things up appwise.
From what I understand, most Linux filesystems defrag silently in the background while the filesystem is being used. Fragmentation can still build up over time, but it takes years of constant use for it to reach any significant level. I use XFS on all my drives, and once in a while I'll run 'xfs_fsr', but not very often.
Whenever I am booting up my Linux, my left click on the mouse is not working and I have to run serval commands to disable and enable device again. Same things happening with my Wi-Fi - it's just spontaneously not working and when I am trying to open "network", my desktop environment crashing and I have to reboot the whole system. I am using Mint 19.3 and cinnamon DE. I am so close to give up Linux and return to the boring Windows OS.
Timeshift just refuses to run, if your target drive has less free space than your entire data on the drive that you try to backup. If you have like 500GB SDD as a single drive for system + timeshift backups and it's halfway full, you are out of luck (even though the partial backup might take only 1% or even less of the free space; I mean like if you didn't even select 100% everything, but like one folder that contains like 10% of your entire data on the drive ... and it's already backed up and the new backup is going to be incremental... it just refuses to run). Reason I stopped using Timeshift.
I don't use Bleachbit anymore, after some time using it, my system started acting weird, now my system works very fine, so sorry bleachbit I'll pass! Also FYI "apt autoremove" removes orphaned packages or packages that aren't needed by any programs installed on your system, just clarifying that info!
Here is 1 for windows10 users go to start search Run THEN type Ping do this several times to Ping your server Ping several times the more the faster. then go to start search Run use %temp% TO CLEAN YOUR TEMP FILES SELECT ALL THEN DELETE THERE YA GO CHRIS GLAD TO HELP lol ops a caps or just copy and paste if you need to so you don't haft to type it lol .
Chris... I'm running MX-Linux on a DELL Latitude 6520. In this Video I was following you in the terminal to update Repositories, I did the "sudo pacman - Syu" command, Terminal said pacman command not found..! I then went to Synaptic Package Manager and looked for it, it's not installed but all files referred to "PACMAN" the Game.. where your pacman eats all the dots...! Later in the video you mentioned pacman, Package Manager..? I tried that and it doesn't work either, so what can I use...? Thanks in advance..! =)
Ubuntu users FYI:
You can freeze any package including kernel packages by:
example-1:
sudo apt-mark hold
example-2:
sudo apt-mark hold linux-image-generic linux-headers-generic
Same with Vanilla Debian ;). Ohhh and don't use PPAs in Vanilla Debian, they will break you install,= only use PPAs with Ubuntu and it's forks.
@@PenguinRevolution can you use PPAs in Mint?
@@matuzaato Linux Mint is a direct fork of Ubuntu. So yes you can use ppas in Mint.
@@PenguinRevolution Are you referring to Arch Linux 32?
@@stephenhunter70 Never mentioned Arch Linux 32. I was talking about debian and Ubuntu
Quick tip: When you type a command that should have been done with 'sudo' but you forgot it, you can type:
sudo !!
This will rerun the last command given under sudo without having to retype the entire command.
when i run linux i run as root lol sudo whats that but i also tend to run it as a live session i have learnt sudo su
though
That's pretty sweet! thank you! I didn't know about this.
"sudo -s" automatically puts sudo in front of the commands you run. You can just type exit to stop doing that.
I'm newish to linux. "Learn to sudo" is basically just putting sudo in front of the commands you're trying to run correct?
@@ghost101049 sudo basically means run as administrator what it does it give you the access to mess with some system files
with Bleachbit never run free hard drive space on an SSD! You'll unnecessarily write zeros to your drive, thus wearing it out faster. On the other hand it's extremely handy to do that now and then on mechanical drives.
Yes never do that on an SSD I prefer Stacey as it is more pleasant to look at and it's more friendly and safer
@@marconyewtv4477 Stacer, not Stacey. right? Genuinely asking as I might install Linux on a laptop soon.
I've been running Ubuntu on my main laptop, which I use intensively, for 5 years now. Haven't done anything to it and works perfectly
switch to mint, thank me later , yes its "ubuntu" minus the input lag
Because of LTT i can't stop reading you channel as "Chris Tech Tips" or "Titus Tech Tips"
LTT = RUclips brand masters! lol.
At least Chris doesn't drop everything he touches like Linus.
@@corey8704 Nice. 👍
Literally why im here
"Titus Tech Tips" +2
then we can call him "Triple T"
Ah, you forgot to add that on Arch-based system we also need to:
- clean pacman cache (all downloaded packages are stored there and taking space)
- uninstalled orphaned packages or packages that were removed from the repo (and we cannot do that blindly, because it often happens that packages are marked as orphaned mistakingly)
- update our configs with pacnew files, this is a manual intervention that is needed to be done once in a while
- update git packages if you have any - they rarely show as an update and this is usually a manual action, although there are commands that can mass update git packages
And updating ur mirrorlist.
One correction: 'apt-get autoremove' removes packages that are no longer needed i.e. packages that were installed as dependencies for another packages that were removed and these packages were left of and are no longer needed. On Arch there is no such command. You'd have to combine 2 commands: 'pacman -Qdtq' which gives you the packages that are no longer needed and remove them with 'pacman -Rsn' or combine them into 1 command: 'pacman -Rsn $(pacman -Qdtq)'.
Also if you use Arch it's probably a good idea to remove packages with 'pacman -Rsn' which removes packages that were installed as dependencies for the package you remove and are not needed by another package and also remove their config files in /etc, instead of 'pacman -R'.
I would say, be careful with what you remove with pacman -Rsn as it is entirely possible to accidentally get cross-dependencies and end up removing something that you actually NEED on your system.
Source: I've done this by installing a Gnome tool, not thinking about it before issuing the command, and wiped out my entire DE including my network manager, so I had to fix it from a chroot.
If you use the APT package manager, I would recommend using nala to upgrade your packages. It's meant to be more user friendly, has a cleaner terminal output, and automatically does certain things for you (autoremoves/autopurges files that are dependencies no longer needed, automatically updates the source list before upgrading so you only need to run sudo nala upgrade, etc.). You also aren't replacing apt, so you can still go in and run maintenance scripts without needing to rewrite them first.
Of course, the apt command family is going to be your de-facto option for more advanced operations, but with nala upgrading is simple and enjoyable.
Chris Titus Tech: Chris. Just a tip. I was always told to never stop bleachbit when it is in the process of cleaning. That you must let it finish or you will have problems.
Note the difference between 'update' and 'upgrade'.
Arch is great. Just type “yay” and you’re done.
Thank you very much Chris, very handy for all new people in the Linux world. Nice video. Have a nice day!
That's really neat. I didn't know that you can keep Linux running indefinitely without a reboot. It's kinda ingrained into me that a computer is going to need a reboot eventually, because of memory leaking or whatever. That's mighty impressive!
Yeah I hear you! I still reboot often because I change tons of things, but yeah... It is not unheard of that many closed systems like elevators, electronic posters, etc... have run linux on a battery backup for 10+ years.
I like when ubuntu requires reboots! Not my
Yea, I know it can be run indefinitely but I still shut down my computer every evening to control my electricity costs, just like I turn off all lights when I go to sleep.
I’ve had macOS with an uptime of nearly three months as well. Memory leaks can happen in software, but especially in the OS it should only be in corner cases. If it’s not in the OS, the OS should be able to clean it up when the offending app is quit.
@@ChrisTitusTech I've just started using Ubuntu and I'm new to Linux. Love your channel. I've noticed that every time I reboot, especially shutdown, I get the black screen and there's a lot of broken code running across the screen. Then it shuts down or reboots. Goes by really fast and bothers the crap out of me. Is this an issue with the bootloader? Any ideas?
I love conky. It's one of my favorite add on apps. I love configuring it lol. It can be an insanely useful tool for monitoring your system...and it looks cool.
3:23
I see Your transparency effects are grainy. I suggest changing composer from OpenGL 3 to OpenGL 2. It helped in my case.
Best wishes ;)
Thanks for the tip Tom!
I find it easier to manage the repositories from synaptic (for a debian system). Just pull down the menu and tick/untick the repositories you have, you can also add new repositories and delete old ones. If you want to remove a package it will show you what else you would need to remove at the same time. There's really no need to edit /etc/apt/sources.list directly.
I also almost never reboot after an upgrade, if there are new graphics drivers, I tend to just kill Xorg and that does the job.
I don't recommend doing most of these except verifying your backups, installing updates, and removing no longer needed and orphaned packages (dependencies of older version of packages that are now no longer needed).
Especially that thing you do with Bleachbit or similar tool is going to cause more problems than it solves. Unless you have a particular problem with a specific application, you generally shouldn't micromanage these cache files. It won't make your system any faster, and more likely than not it will just slow down programs and cause data corruptions.
In low disk system, it may be worthwhile to clean the package managers' download caches, but if you aren't running low on disk space, you don't need to worry about it.
One maintenance that I recommend that hasn't been mentioned, if you're using Ubuntu-based system on btrfs system, you may need to remove the automatically created btrfs snapshot that apt created before a major system upgrade. Once you've confirmed the upgrade is successful and you decided you likely won't be rolling back, you can just delete the snapshot using btrfs tools. apt won't remove them automatically, and these snapshots take up quite a lot of disk space (and because of copy on write, they'll take even more space the longer you keep them).
Yeah once in a while I'll just see how much storage space I have left. Linux can get loopy if you get up to 90% full. Up to then no worries though. Right now my root partition is at 32% so no problem there. If I get up around 80% I'll think about maybe ditching something then. Or I'll just get more storage space. Probably the latter knowing me.
Subbed. Just starting my linux journey. Years late I might add. Lol. I want to move my family over from windows and Microsoft.
Greetings, I have been transitioning to Linux exclusively for a few months now. This video has been very helpful. This and your other videos have been very helpful. Ciao!
Great tips! Definitely love information like this for cleaning out old junk that builds up over time.
Stacer is a really nice system cleaner worth trying out. Then there is Sweeper, Deborphan and Debfoster
You do the best tutorials Chris! Clear instructions, reasons why you're doing stuff, great explaintions and poiniant tips. But best of all is your smooth delivery and non-intimdatng predisposition. Love your work mate, it gets me pumped about the future of the world. Thank you kind sir. 👍💯
Many of my users are running my software on Win 2000 still. No maintenance required, but using very few Windows facilities. POS/Inventory and such. Linux is better, but less familiar and has fewer available third party programs. In Windows, I wrote my own scheduler as I did not trust the one that came with it. This can be done. There is push back on Linux desktops. Linux servers are well received, is my experience.
WHAT!?! Zero playtime on World of Warcraft? Have you noticed, the WoW Classic has just been released?
That is an amazing UI, and configuration that you have going on there, any insight to what all you have installed on there for that?
On the linuxmint forums, they warn that bleachbit is bad, that it can break your system, and that you shouldn't use it. I have used it for a while and I have had no problems.
I’ve been watching your videos recently and they are awesome!! Keep up the good work!!
Thanks Chris for warning users about doing a backup before purging packages, purging packages is a very dangerous operation, you can mess up your system beyond recognition.
6:18 and 12:43 when you forget "sudo" in front of a command, you can look 110% more cool if you run "sudo !!". It's a proven scientific fact.
@callam85 Ladies like men who use "sudo !!". You're just jealous.
callam85 depends on the shell. Standard bash logs “sudo !!” as the full command string
Just to expand on the no need "no need to defragment" part on linux.
The reason linux and most unix/bsd filesystems do "not need" defragmenting is because they leave more space between files to allow them to grow on the block level unlike ntfs which uses the immediate space (pretty much) after a file is done being written so any newer data for the file has to be written elsewere and a reference pointer is added to the end of the old data ("continues at XYZ") rinse and repeat, however you can run into situations where defragmenting the filesystem will be needed, however it is done in the background and a manual run might only be needed when the disk is close to being almost full (which in most cases will be rare)
so yes you dont need to defragment your system like you have to do on windows unless your disk is pretty full before you will need to do a manual run (we talking maybe 80-90% full)
before that, all defragmenting is done as needed in the background without you noticing any impact from it but you have to go over 50% disk space used by average before defragmenting is even done in many cases, as no files will be fragmented (for the most part) before that limit is reached.
HikariKnight keep in mind that SSDs should never, ever be defragged. Not only does it waste lifespan, but if the controller is cheap you can brick the drive.
@@jadosrd8950 If this ["brick the drive"] would be true, any IO access pattern could brick your drive: You should just not (and don't want to) use such drives. There are applications where "defragging" an SSD can make sense but that needs special defraggers that care about reducing IOPS for your specialized application and not moving files around as most classical defraggers do (moving files around is pointless on SSD, and introduces additional write load). But "never ever" is also: Your SSD wouldn't just die the next day just because you defragged it yesterday. It lost maybe 0.5% of it's life span, on modern drives probably even less. The bigger the drive, the less prone it is to this problem: Bigger SSDs have a exponentially longer live span for non-suitable write patterns.
Kai Krakow again back to the controller chip. A good, or even half decent, chip can withstand defragging, but it also makes it a moot point as THAT IS it’s JOB to read/write a bunch of scattered data as if it were in one chunk. A bad controller, and Im mainly talking $10 Ali junk here, has been shown multiple times to not “keep up” with knowing where data is and subsequently “losing” important data (Windows files).
@@jadosrd8950 Yep, that's probably the type of controller you don't want to trust your data - no matter if you defragment or not.
Thank you, Chris. I love apt.
This video is missing out on handling RPM-based systems like openSUSE.
old habits die hard, no reboots neccesary :p
just need to figure out what needs to be restarted xD
Great video, really useful information on keeping Linux running smoothly. Thanks Chris
bleachbit is an underrated lifesaver
Makes me wonder why Microsoft just doesn't create their own distro like RedHat. I mean the biggest thing is software, but if they created it alongside windows, then use something like a .NET virtualized container, while supporting developers that create main ports to their vision of a linux OS. I mean powershell and windows server is a huge hurdle to overcome, if they were to do that, because powershell is object oriented and linux is textual. Windows Server has implementations all over enterprise environments. So actually fundamentally changing Windows may be too volatile for them.
I like your videos, man! Based on your teachings, I finally decided to get rid of Windows ...
Hey, nice video, man, but I saw you play minecraft over lutris. Minceraft has native Linux suppor, you should try it
"FAT is event worse...". MS has just open sourced the exFAX spec and is supporting exFAT being added to the Linux Kernel. Let's see if MS will pull this off. I think they will.
Why would we want it! And why add it again?
@@stephenhunter70 it's used on many external drives. Nobody is using it for the system drive even in Windows.
@@Alex11223l2 Exactly, I rest my case.
@@stephenhunter70 my external SSD uses it, so I can't use it on my Pi.
It's gonna add support so they are not going to switch the linux file system. You can just use it without formats.
BleachBit is an excellent program. I didn't know I could do a sudo bleachbit though. Thanks Chris.
Awesome video man, this is the kind of stuff I'd love to refer to people who are just getting started with Linux. I would also add `htop` to the list too as for me it stands in place of windows task manager. Love the channel. :)
I find that top is a bit easier to uses and easier for a new guy to figure out.
I'm currently doing a personal experiment with my Manjaro KDE system. After a couple of update breaking (that i restored with Timeshift), i decided to completly stop doing any updates whatsoever. That was back in May of this year. So far everything i have installed back then work fine. I did install one or two other programs that work fine, but some others i could not because the dependancies are not up to date. I have another system on another SSD, Pop Os, that is up to date and can utilize for whatever reason. But in conclusion, my Manjaro still works like a charm and i use it every day as my main system!
allmost weekend, nice video chris....
In addition to `apt autoremove` I do recommend running `apt-get auto-clean`. It looks like `apt` likes to hoard every downloaded package and that could easily get over gigabytes. I had switched to 'testing' that had regular updates. It has gotten around 8GB of cache... on my 32GB sdcard. To empty the whole cache you can do `apt-get clean`, this however would remove the archives of the currently installed packages too.
Great video! Thanks for the tips and tricks Chris!! Nicely Done!!
Oh come on, Chris... I'm a fan of Linux, too. And I don't like Windows for a lot of the same reasons as you. But saying, that NTFS is just a slightly improved FAT file system is just technically wrong. In fact, NTFS is an extent-based FS just like ext4, and it has a journal similar to ext4, maybe better compared to XFS. And what's also not true is that defragmentation is needed in NTFS and not in "Linux". In fact, neither needs defragmentation, both do - thanks to extents - a pretty good job of keeping fragmentation low, as long as you don't fill your drive too much or have some very strange file access patterns. Classical defragmenters are mostly useless. What should be done, tho, is using an advanced optimizer which puts your files back into a better location (i.e. "MyDefrag" can do this, and even the Windows-integrated defragger does this in some manner). Why is it needed? Because Windows Update messes with the optimal placement of your DLLs a lot, this slows booting and application starts a lot over time. This is not a fragmentation issue, it's a file placement issue. And guess what: Linux suffers the same problem, constant updates mess a lot with file placement over time. This is why there exist boot-time read-ahead programs, and boot-time file location optimizers (i.e., I think e4defrag can do this). BTW: File location optimization is not needed on SSDs, just do not care about this when you're on SSD.
But still, Windows can suck a lot more when it comes to hard disk IO, and this is mostly not due to fragmentation. This is because Linux has quite a good IO scheduler, while Windows mostly doesn't (it's a very simple one and doesn't care much about fairness, it just looks at which processes are interactive and which are not, I think). In fact, MS thinks this is not a problem the OS should solve because you can either use SSDs or advanced RAID controllers that do it for you. But again, this is not a FAT vs. NTFS vs. Linux file systems problem, it's an issue baked (or not-so-baked) into the OS itself.
Yea, the IO scheduler in MS Windows is in desperate need of an overhaul.
Since anti-viruses just seem to take top priority... which just seems to kill performance...
THANK YOU. NTFS has been around a long ass time and it's been far and away better than FAT32 for a long ass time. This seems to be a common theme among Linux users. They think the FS Windows uses is this slow outdated file system when it's actually incredibly competent and has been steadily updated for years now.
@@arnox4554
The drivers available for Linux to mount NTFS partitions are terrible and combined with the IO issue in Windows. It creates the perception to Linux users that overall, NTFS is terrible.
The NTFS drivers are terrible as they have to essentially be created with reverse engineering and little to no external or, it seems, internal documentation is available to the public or appears to exist at all on how exactly the NTFS works under the hood.
@@hanro50 Yeah, that's what I was suspecting too. I think ext4 also suffers though under Windows? I'm not sure, but it really looks like if it's not native ext4 or native NTFS, the performance goes down the toilet.
@@arnox4554 I think it comes down to a simple fact: The IO layer of Windows is designed much around the design of NTFS, the single file system of Windows (in terms of "as the system drive"). That may have some advantages compared to Linux which has a lot of filesystems. So it's probably expected that any non-native filesystem will perform exceptionally bad in a lot of corner cases. But that may also be true for native filesystems, i.e. I'm using btrfs and I like it a lot for different reasons. It's not exceptionally fast because it's copy-on-write. This is not my issue, you can observe the same problem with ReFS in Windows. But let's look at btrfs integration: It's not a traditional block-based file system as, e.g., ext4, it behaves more like NFS (yes, not NTFS, I mean the network file system). Why is that? Well, there's not a direct mapping from the block device to the FS, that makes it a virtual file system with virtual block devices: If you mount a btrfs, the device file acts more like a network address than a block device because btrfs supports sub-volumes and multiple devices. That means, as for any virtual file system (aka not mapped to a block device) that we have unstable st_dev and st_inode members in the stat syscall: Across reboots, the returned values may change. It's only stable for the duration of the mount. This can confuse a lot of software, i.e. baloo in KDE or backup programs that detect differences by values returned from stat syscall. What could we say? The virtual filesystem layer in Linux wasn't designed around the idea of having a non-1:1 mapping between block devices and filesystem. That's the kind of design I'm talking about. I'm pretty sure the IO scheduler of Windows has a lot of assumptions and design choices build around NTFS implementation details. That allows it to be much simpler while at the same time being not too bad. But there's room for improvements on both sides.
So it's not a question of which FS is better, it's a question of integration and design. NTFS was made for Windows, it's not a POSIX filesystem. Ext4 was not made for Windows, it's made for the Linux VFS design. I found NTFS to be incredibly stable in Windows (given the right underlying storage layer), be it performance or reliability. I cannot actually say the same for ext4 in Linux (tho, my experience is some years old for this), I found XFS to perform way better. Also, NTFS has gained a lot of features that Linux filesystems are still missing: Online hot spot repair (repaired while running without requiring a full chkdsk scan), offline hot spot repair (for what can't be fixed online but also without full scan so it's fast and silent), volume snapshots... Btrfs tries to do something similar but approaches the problem from a different direction (similar to ZFS). ReFS has integrated multi-tier support (you could achieve similar results with bcache + btrfs, and bcachefs is on the way to integrate something similar). There's a lot of exciting stuff going on currently, and in this regard, ext4 is just from the stone-ages (but also "rock solid", haha, stone, rock, you get it?), it hasn't gained the same amount of feature improvements than NTFS although they seem to be based off the same ideas and were designed at around the same time, I think.
So, let's conclude: What is good and what is crap is mainly a question of expectations. Even fat32 can be totally perfect for your application. But almost always something may become crappy if used in the wrong way. From this perspective: ext4 is quite crappy for most of my applications.
Two things I noted here.
1. As a general rule don't pin packages on Arch. Partial upgrades are unsupported and are prone to causing system breakage.
2. When removing packages on Arch use sudo pacman -Rns instead of just -R option. That will also remove any "orphaned" dependency packages and leave you system much cleaner.
Edit: Rephrased for clarity.
I love your videos, but as a brand new user, it would help me if you slowed down just a little bit and included a few explanations of some of your terms. Otherwise, you have been a lifesaver for me!
Bleachbit is too aggressive and too complex for an average user. I thought I know what I was doing but after some time I discovered some small breakages, lack of data that most probably was "cleaned" by bleachbit and I was very, very careful and set only most general and basic cleanup and yet it was still too invasive.
@BlackWorm trying to find that but don't see it. Can you please share the link mate?
In case you have packages that can be upgraded you can run the following:
'sudo apt upgrade' or 'sudo apt dist-upgrade'
Cheers!
Alright. Needed this. Nice Video Chris
In my experience, "sudo bleachbit" and "sudo rm -rf /" are similar in effect.
Hey Chris, nice video as always! Could you make a more thorough arch install video sometime in the future? :)
@@rdangdev UEFI
Since windows 93, people have complained about buggy op systems. It looks like if as much effort was put into Linux, it would be perfect by now.
I am not very fond of command-line. I am not afraid of it, I just use GUIs whenever available. Call me lazy if you like. I used Mint Software Manager to install Bleachbit. So, let's continue with the video.
exactly what I needed, thank you very much man! :D
Chris you should never, ever hold or freeze packages or do partial updates in Arch Linux, madness lies that way, its like feeding a Mogwai after midnight, just don't. one update tip I have is to install reflector and set up a service
###
[Unit]
Description=Pacman mirrorlist update
[Service]
Type=oneshot
ExecStart=/usr/bin/reflector --age 24 --protocol https --latest 10 --sort rate --save /etc/pacman.d/mirrorlist
###
Then set up a bash function in ~/.bashrc
###
function UD {
sudo systemctl start reflector
sudo pacman -Syu
#yay
}
###
then save and
source ~/.bashrc
then just type UD in konsole to get the most recent updates from the fastest source repo
....or if a fellow KDE user install discover, appstream-qt & archlinux-appstream-data (and deps) so KDE notifies you of and can apply updates for you.
Depends on what your holding back.
@@PenguinRevolution not really, one of the main provisos of using Arch is no partial updates/ held packages, do so and at some point you will regret it. Arch rolls and packages can move from core to extra frrom extra to community to AUR and from AUR to orphan, and vice versa, without warning or notice, if you have frozen a package and that happens to a dependent or dependency then....
Check out the arch wiki pacman entry and search for 'partial', while pacman.conf has the 'ignorepkg' and 'ignoregroup' sections its probably not wise to use them unless you are using a private repo you control the packages within and know they have no dependencies outside of that private repo.
@@mrmysteryguest I understand the concept of rolling release, so you don't need to educate me on that. I stand by what I said, it depends on what you are holding. You can get a way with a lot if you know what you are doing. You can hold your kernel without any consequences, there was a time when that was a requirement because updating kernels could break your system. also if you holding something like GIMP or LibreOffice then you probably can get away with it also. If something breaks all you normally have to do is un-hold your package and update it. I've used Arch in the past and I've held packages back and never had a problem.It's also potential problems like this is why I say running arch is never a good idea anyway, at least not for most people.
@@PenguinRevolution if with the last third of your post you are saying, "I used to run Arch, I held some packages, updated, had problems then stopped using it and advise others its not stable and breaks a lot', I refer you to my original post about feeding Mogwai's
Also historically holding GIMP was not a good idea due to its connection to the GTK Libs
If there is an upstream issue with a package like a kernel, the idea is to carry out any instructions on the Arch news page - eg to use the live media to chroot into a broken install to roll back using the pacman cache while its fixed upstream (extreme example), I've known this to happen but this is not the same as holding or pinning a package in Arch, which always brings issues in the longer run.
Most ex users I know who had issues with rolling distros did so because they stepped outside of the the distro's recommended processes often by carrying out partial updates or pinning packages (thats not just Arch)
@@mrmysteryguest Arch is unstable in nature. everybody has breakage at sometime. THat's why it's never a good idea to run arch as your daily driver. In my opinion Arch is an unstable mess.
You should also keep an eye on cron. Make sure nothing got put into your (or root's) crontab and remove old cron jobs that you don't need anymore.
Ironically one of the things I learned from this video is the package name for kcharselect. I really badly needed a program like this and Kubuntu didn't ship with it. Also tried bleachbit and got 20GB of space back, which is great :3
im relativity new to the linux world , but i believe my pc experience carries through .. for clarification i remember when Microsoft released there first operating system and have used cpu's measured in hertz not kilohertz or even gigahertz! and ram was measured in k's so i believe that dates my usage experience. from my experience all os;s degrade with user input.. all os;s have recommended maintenance.. all will sorta fix the issues and take time... installation of every os ive ever done takes around an hour solving cleaning and troubleshooting always takes longer,, having an archive of all drivers and software installed says if data is saved to non stand locations a full install of the os and software is less effort and time than "maintenance" ever was and has far better results for me anyways mileage may vary
Chris I believe Fedora uses DNF package manager instead of YUM, REL/CentOS still uses YUM for the moment but with REL 8 I believe they will ship it with DNF. Also to complete the update you need to run sudo apt upgrade.
Penguin Revolution Yeah I was about to mention the same thing
@@JohnMiller-mmuldoor A lot of people get that mixed up these days (I do it sometimes too) The only reason REL doesn't use DNF is because Red hat hasn't had a major feature update since DNF was released. and REL and CentOS also provide you with a path to replace YUM with DNF if you so choose, however most people don't and in my opinion if you do use REL or CentOS then it's not worth switching it out until the next major release comes out, YUM works just fine.
There is defrag for ext4 and xfs. Run e4defrag -c /location/ to check your non ssd drive for fragmentation.
Wow its so much simpler compared to a "how to debloat windows 10". Who knew arch was simpler than win10?
Timeshift rescued my main Linux install yesterday...
Do people like me need time shift? I just browse the internet and occasionally game
/tmp is not cleaned by the system. Linux does not delete your files unless specifically told to. The impression the /tmp gets cleaned upon reboot comes from the fact that on most distribution a special filesystem called tmpfs is being mounted at that location. tmpfs is a special filesystem that lives in RAM rather than on disk - that's why it gets cleaned when the system shuts down. Linux has nothing to do with it. If you tell it to mount a disk partition at /tmp or not to mount anything there at all, it won't and your temp files will then survive the reboot.
Had to clean the cache every couple months or the root partition would fill up and I couldn't update my Arch. Started to annoy me, so decided to mess around in package manager and keep less copies/versions. Now it''s fine.
I had to fsck? After sometime of use, I told followed info i found on the net, can you explain what that is and how i can not have to do it in future
BLESS YOU CHRIS
For Fedora 31 you will want to run the following in the terminal prior to launching bleachbit once installed:
sudo dnf install python2-gobject -y
It works fine once you do this. This is from their News segment on their site (link: www.bleachbit.org/news/bleachbit-30). Enjoy!
Very interesting and informative.
Thank you very much for the video!
I don't know anything about btrfs, zfs and HAMMER, but it appears to me that there is some performance increase after defragmentation a ext4 volume. Ubuntu even comes with e4defrag utility preinstalled. Here is the article I am referring to:
www.linuxsecrets.com/kdocs/ols/2007/ols2007v2-pages-179-186.pdf
Many People duel boot so they will be using NTFS.
Timeshift is not compatible with Linux Ubuntu MATE, at least not on my PC because it screwed up my system and broke it when I did a restore.
Some Linux hardware h has to be defragged. SSDs and M.2 drives need periodic trimming. This is usually scheduled by Linux on a regular basis such as "every 200 hrs of use, or after xxx gigs have been written. Same requirement for Windows systems.
Eventually, the SSD bios will have an extension to where the SSD itself, without Linux, will do its own trim.
I like to run straight Debian. maintenance is via the apt command and this seems simple enough
I use Discover and Synaptic Package Manager to uninstall broken PPA and residual config. Also with the autoclean, autoremove commands. However I'm reluctant to use Bleachbit, had problem with it in the past.
I guess I'm just lazy because I don't do any of this. I just run aptitude update;aptitude safe-upgrade occasionally. As long as I'm not running out of disk space I never bother uninstalling anything. I'll make sure nonsense processes aren't running. But I could care less what's wasting disk space. I got plenty of that. So maybe you can do a video about how you examine running processes? pstree is cool.
NTFS is not crappy, pretty modern FS with few drawbacks. With proper usage, it's perfect for specific tasks. I have done a ton of research on file systems, and ext4 also has lots of weaknesses, for example inode allocation, no cow functionality etc. Every major Linux FS has some kind of defrag... its just how rotational media works, after some degree not much you can do to avoid it. I keep on my eye on SGI's XFS and Btrfs, in the future they have the potential to be the next big hit under Linux.
XFS is already a great filesystem. It's what I've used the majority of my 10+ years of running Linux as my main OS. Even NASA uses XFS.
@@adamn.4111 They working on XFS copy on write support and snapshots.
Krisztián Kovács ntfs is nowhere near modern (1993). It was designed for low latency microwrites on striped arrays in NT Server 3. It just happened to be more modern and more stable than FAT. Not saying it’s “bad,” but the only reason it made it to consumer Windows was lack of a better option for large drive support
@@jadosrd8950 I'am talking about NTFS 3.1 (NTFS 5.1)
It's not fair to compare ext4 with ntfs 1.0 from 1993. But if you compare NTFS 1.0 with Ext or Ext2 from the era it's not that terrible. In the SSD age no one cares a about fragmentation, and as is said before ext2 was also terrible. In 1993 many ways inferior compared to NTFS.
Chris, what did you use as FPS-counter while running Minecraft?
Peter, I just used the built-in game FPS counter.
@@ChrisTitusTech
I never played Minecraft.
have been updating arch everyday without reading shit in the last 3 months, i'm surprised i didn't break it...yet :D
Great video! I'm sure you may have said before, but I'd like to know what desktop environment you're using on this machine. It looks great.
The Arch machine, to clarify.
Just saw it in another video. KDE looks slick.👍
Yes you dont need to defrag. But it can be good on a normal disk
Linux systems can get fragmented if you have very little disk space, but otherwise it almost never happens in ext4 (which is the defacto fs on most distros)
Well it still happens with a lot of disk space actually. It’s just that it doesn’t happen for long, because when the disk isn’t super active it’ll reallocate things (pretty quickly) to fix it itself
apt autoremove in some case may remove critical stuff. Its basically all packages with no deps.
Hey Chris. Just noticed that this video is from 2019. Can you make an updated version of this? Like maybe there's something new or noteworthy that you can show us in 2020. Thanks!
should do a: when i upgrade to a new major change distro version. Something i noticed when upgrading from pop os 19 to 20 it broke alot of app installs. i use alot of different tools to install: pop shop, apt, flatpak, snap, chocolatey. it can really mess things up appwise.
I have a problem that makes me reboot, which's random sudden lasting CPU spikes, once CPU goes crazy, only way to fix's a reboot.
Fragmentation/defragmentation is possible on ext filesystem ;)
But usually not required
From what I understand, most Linux filesystems defrag silently in the background while the filesystem is being used. Fragmentation can still build up over time, but it takes years of constant use for it to reach any significant level. I use XFS on all my drives, and once in a while I'll run 'xfs_fsr', but not very often.
Whenever I am booting up my Linux, my left click on the mouse is not working and I have to run serval commands to disable and enable device again. Same things happening with my Wi-Fi - it's just spontaneously not working and when I am trying to open "network", my desktop environment crashing and I have to reboot the whole system. I am using Mint 19.3 and cinnamon DE.
I am so close to give up Linux and return to the boring Windows OS.
Timeshift just refuses to run, if your target drive has less free space than your entire data on the drive that you try to backup. If you have like 500GB SDD as a single drive for system + timeshift backups and it's halfway full, you are out of luck (even though the partial backup might take only 1% or even less of the free space; I mean like if you didn't even select 100% everything, but like one folder that contains like 10% of your entire data on the drive ... and it's already backed up and the new backup is going to be incremental... it just refuses to run). Reason I stopped using Timeshift.
I don't use Bleachbit anymore, after some time using it, my system started acting weird, now my system works very fine, so sorry bleachbit I'll pass!
Also FYI "apt autoremove" removes orphaned packages or packages that aren't needed by any programs installed on your system, just clarifying that info!
Hi Chris!
What do you think about coreboot and its variants? Have you ever tried installing it?
This is very helpful
I thought Conky was Bubbles' puppet's name
I never did maintence in windows... It is slow...
Edit: I did maintence... It is still slow...
Here is 1 for windows10 users go to start search Run THEN type Ping do this several times to Ping your server Ping several times the more the faster. then go to start search Run use %temp% TO CLEAN YOUR TEMP FILES SELECT ALL THEN DELETE THERE YA GO CHRIS GLAD TO HELP lol ops a caps or just copy and paste if you need to so you don't haft to type it lol .
Chris, how did you reboot your machine whilst recording the video?
That cute smile
It’s old but I’ll add my 2 cents. Also try removing orphan packages from time to time
uCareSystem is great tool for Debian based systems.
Chris... I'm running MX-Linux on a DELL Latitude 6520. In this Video I was following you in the terminal to update Repositories, I did the "sudo pacman - Syu" command, Terminal said pacman command not found..! I then went to Synaptic Package Manager and looked for it, it's not installed but all files referred to "PACMAN" the Game.. where your pacman eats all the dots...! Later in the video you mentioned pacman, Package Manager..? I tried that and it doesn't work either, so what can I use...? Thanks in advance..! =)