Linux - GeneralThis Linux forum is for general Linux questions and discussion.
If it is Linux Related and doesn't seem to fit in any other forum then this is the place.
Notices
Welcome to LinuxQuestions.org, a friendly and active Linux Community.
You are currently viewing LQ as a guest. By joining our community you will have the ability to post topics, receive our newsletter, use the advanced search, subscribe to threads and access many other special features. Registration is quick, simple and absolutely free. Join our community today!
Note that registered members see fewer ads, and ContentLink is completely disabled once you log in.
If you have any problems with the registration process or your account login, please contact us. If you need to reset your password, click here.
Having a problem logging in? Please visit this page to clear all LQ-related cookies.
Get a virtual cloud desktop with the Linux distro that you want in less than five minutes with Shells! With over 10 pre-installed distros to choose from, the worry-free installation life is here! Whether you are a digital nomad or just looking for flexibility, Shells can put your Linux machine on the device that you want to use.
Exclusive for LQ members, get up to 45% off per month. Click here for more info.
View Poll Results: Which Is Your Preferred Linux File System?
jfs... because ibm is god and re-sizing your file system is heresy
JFS is also mature technology just as I and my systems are mature. Some day EXT4 will grow up and emerge from beta-testing but until then I will choose JFS for systems that matter.
Of particular note, ext4 is only used by people because it is the default. If another filesystem were the default people would soon forget about ext4. I honestly think ext4 is the filesystem that is outdated and should be replaced. It uses a B-tree, whereas modern filesystems use a B+ tree (like NTFS, ReiserFS, NSS, XFS, JFS, ReFS, and BFS). https://en.wikipedia.org/wiki/B%2B_tree
There have also been a number of issues with the ext4 writeback mode causing data loss, and for a while it was the default mode.
Since we know Linux is for Geeks, and there's probably a reason but for us real worlders what's the difference and why should we care. Is this a coresponding issue for Apple users?
It's not an "issue" for Linux users.
A scan of your posting history shows that most of your earlier posts were similarly scornful, condescending and unconstructive. Not to mention insincere: your "question" is what the replies to this thread were already answering. You've been behaving like this for close to a decade. Please stop.
I tried many FS for years.
Ext4 is good, secure but not fast enough. Btrfs is a pity. Reiser was particularly fast but in the meantime not sure, I lost too many critical files(v3).
Finally, my choice is XFS because, it's fast, quite sure (I suffered from empty files due to a crash, particularly with KDE).
JFS (Jesus For Safety), I don't trust... Sorry, each of us can have its own opinion, noop! :-p (joke).
In fact, I don't have a real preference for a filesystem, it depends a lot on what I require from it, so, I must compose between speed, security and evolution.
Location: Nacka, just outside of Stockholm, Sweden
Distribution: Manjaro Mate, Ubuntu Mate, Debian
Posts: 22
Rep:
I have used most of the Linux file systems during my 20 years as a Linux-only computer user and find the most stable to be JFS though I liked XFS that I ran on a server, that would be 2nd in line for me. I tried BtrFS way too early I think (2010) and was greatful for the backups I took as the HDs were pretty messed up (still don't know what went wrong). Read about it being considered stable as of mid 2013 so I am tempted but still reluctant from my earlier experience. I may wait for encryption to be implemented (or is it already??). Many interesting features, interesting B-tree variant.
skikir: as for the "geek" comment, large server systems real worlders use Linux or Unix, a tiny fraction of the Unix slice use OS X (server version).
As for web servers: 37% Linux, 33% Windows, 30% Unix (whereof a small part OS X but mostly FreeBSD, HP-UX, AIX and Solaris).
Supercomputers use Linux to 97%, 2.6% use other nix:es (not OS X).
As a typical desktop user, I've not found much need for much more than ext4. I think it's a common default and so I like to pretend that counts for more than the alternatives. I'm eyeballing btrfs because it does some Really Neat Things™, but that's it.
I sure wish some of the replies would go into a little more depth on the speed of their file system. I'm aware of benchmarks that do things like create tons of directories and files and then time various interactions which will usually reflect design characteristics of the file system, but I haven't read so many practical reviews of such situations, or other scenarios where one was too slow but a switch to another fs made all the difference. That's just my curiosity; no doubts implied.
I sure wish some of the replies would go into a little more depth on the speed of their file system. I'm aware of benchmarks that do things like create tons of directories and files and then time various interactions which will usually reflect design characteristics of the file system, but I haven't read so many practical reviews of such situations, or other scenarios where one was too slow but a switch to another fs made all the difference. That's just my curiosity; no doubts implied.
For me the difference between xfs and ext is night and day as soon as the file count and size go up. The time required for simple file creation, directory creation, and ESPECIALLY things like du are orders of magnitude apart.
Here's a quick benchmark to give you an idea:
Machine 1
Code:
$ df -T .
Filesystem Type Size Used Avail Use% Mounted on
/dev/sdb1 ext3 11T 3.6T 7.2T 34% /home
$ find . -type f | wc -l
215706
$ find . -type d | wc -l
4999
$ time du -sh
440G .
real 0m57.881s
user 0m0.177s
sys 0m2.177s
215k files, 5k directories, 440 GB, du took nearly a minute on ext3, and in my experience ext4 isn't much different.
Machine 2
Code:
$ df -T .
Filesystem Type Size Used Avail Use% Mounted on
/dev/sda1 xfs 73T 38T 36T 52% /home
$ find . -type f | wc -l
868778
$ find . -type d | wc -l
94783
$ time du -sh
1.2T .
real 0m17.359s
user 0m1.183s
sys 0m10.740s
870k files, 95k directories, 1.2 TB, du took 17 seconds on XFS.
Here's another directory on the same XFS filesystem
Code:
$ find . -type f | wc -l
279560
$ find . -type d | wc -l
5455
$ time du -sh
37T .
real 0m0.448s
user 0m0.039s
sys 0m0.237s
280k files, 5.5k directories (pretty similar to the ext3 directory above), 37 TB, du took less than half a second.
The finds also ran faster on the XFS system, filesystem navigation is faster, searching is faster, etc.
I use ext4 (has worked well) but did not vote because never looked into what should be preferred(, at least for my needs(stability\speeds (all I care about...))
Last edited by jamison20000e; 11-22-2014 at 03:24 PM.
I use ext4 (has worked well) but did not vote because never looked into what should be preferred(, at least for my needs(\speeds (all I care about...))
Maybe the ext4 option should be split into "because it is the default" and "because I choose it over others".
Distribution: Debian Sid AMD64, Raspbian Wheezy, various VMs
Posts: 7,680
Rep:
I use ext4 as it happens to be the default and has not let me down. I tried BTRFS a while back and had problems -- it may have been the drive it was on not the FS but without the option of scientific testing it did seem to be the FS. Things may be better now, I may try it again.
I have to admit ignorance in not knowing anything about jfs but I must take a look now.
Slightly off topic but, from posters here, I have learned of UDF and am planning on trying it as the file system for my unencrypted portable media.
Last edited by 273; 11-21-2014 at 07:31 PM.
Reason: Typo'.
for us real worlders what's the difference and why should we care.
You don't have to care if you don't want; the default ext4 which is currently used by many distros is quite adequate and reliable.
There are some new features in systems like btrfs which I am personally looking forward to, like better fault tolerance, data deduplication through copy on write, and live backups. I'm conservative about not switching to it until the dust has settled, but I think that will be soon.
Distribution: FC-19 as primary, FC-20 in VM. Debian on my NAS.
Posts: 4
Rep:
Started with Berkeley Fast File (UFS or FFS for you fellow old timers) on BSD 4.2. Transitioned through ext2/ext3 (there for many years). Journalling is what made me move to ext3 from ext2 (lost a bunch of filesystems). I was very happy with that in all production systems (with journalling) until volume sizes really started taking off. I/we then moved into ext4 when it stabilized. Been on ext4 since.
My NAS is ext4 on Raid 5. I can pull the drives, put them into enclosures and move to another linux machine if needed. (yes painful but doable)
Code:
Filesystem Type Size Used Avail Use% Mounted on
moesern:/v1/homes/chuck nfs4 26T 15T 11T 59% /chuck
Last edited by ChuckPa; 11-21-2014 at 11:20 PM.
Reason: formatting.
LinuxQuestions.org is looking for people interested in writing
Editorials, Articles, Reviews, and more. If you'd like to contribute
content, let us know.