Announcement

Collapse
No announcement yet.

Benchmarking ZFS On FreeBSD vs. EXT4 & Btrfs On Linux

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Originally posted by jalyst View Post
    opensolaris is alive and well.
    Yeah, right! When will be next release?

    Comment


    • I couldn't care less, find out for yourself.
      I do know that there's no reason for it to disappear of the face of the earth.

      Comment


      • Originally posted by jalyst View Post
        I couldn't care less, find out for yourself.
        I do know that there's no reason for it to disappear of the face of the earth.
        It seems they've got some problems 'thanks' to Oracle. This about Solaris is dead I said to Kebabbert only. ;>

        Comment


        • Originally posted by kebabbert View Post
          Maybe I was not clear enough, or you misinterpreted me. Anyway, as of now, Linux has limitations. But in the future, Linux most probably will only be better. IBM basically says this too. IBM AIX is phased out, it will be killed. IBM will bet on Linux instead. This is outspoken and official. So, basically, IBM also says that Linux is not mature enough right now. IBM sales personal always tries to sell IBM AIX which is very expensive, and as a last resort, IBM tries to sell Linux. But IBM does not try to sell Linux the first thing IBM does.



          I have read on a link here, that there is only one full time developer on BTRFS. I know there are part time developers, and probably some hobby developers working on their spare time. But still, if Oracle is really seriously committed to BTRFS, they should dedicate much more resources, preferably some ZFS developers. You can not develop an Enterprise filesystem in the basement, by some part time developers. If Oracle wants to sell and earn money on BTRFS, then it could be wise to develop most of BTRFS in house? Now there is one full time dev, and other part time developers. I dont see how you can get an enterprise product to support and sell, from such a small effort?

          But maybe I am wrong, as you say. Or maybe I am right: now that Oracle has ZFS, then Oracle quickly lost interest in developing another filesystem that does exactly the same thing as ZFS does, without adding something new. Why develop two identical products? Does not make sense from a business perspective. But time will tell. Maybe both of us are wrong!
          Definitively you are wrong.

          Oracle uses linux and has linux customers interested in datacenters. Oracle itself uses linux internally. So yes, oracle is very interested in btrfs.

          Btrfs is a true open source project. it not depends entirely on Oracle, how many plp work at linux kernel (Linux International)?? and how big and evolved is linux kernel?? Btrfs project is similar, lots of devs from different companies: Red Hat, SUSE, Intel, IBM, HP, Fujitsu, etc are working on it because a lot of linux customers and companies are interested. Please stop FUD on this and read:

          http://www.mail-archive.com/linux-bt...er.kernel.org/

          Comment


          • Read here too:

            git.kernel.org/?p=linux/kernel/git/mason/btrfs-unstable.git;a=shortlog

            Comment


            • Originally posted by Jimbo View Post
              Definitively you are wrong.

              Oracle uses linux and has linux customers interested in datacenters. Oracle itself uses linux internally. So yes, oracle is very interested in btrfs.

              Btrfs is a true open source project. it not depends entirely on Oracle, how many plp work at linux kernel (Linux International)?? and how big and evolved is linux kernel?? Btrfs project is similar, lots of devs from different companies: Red Hat, SUSE, Intel, IBM, HP, Fujitsu, etc are working on it because a lot of linux customers and companies are interested. Please stop FUD on this and read:

              http://www.mail-archive.com/linux-bt...er.kernel.org/
              I am only saying that I believe if Oracle is serious about BTRFS, they should have dedicated far more resources internally than one full time paid developer. That is not FUD. It is true. And I do make it clear it is my BELIEF. I do not have proof or evidence that Oracle is abandoning BTRFS. It is just a guess from my side.

              And I also say that normally, a company does not develop two identical products. And many people agrees that ZFS is superior to BTRFS. So I am asking, why develop an inferior copy, instead of betting on the original ZFS?

              I make it clear that I guess and I think so. I do not claim to have it as facts. I know there are more than one full time paid developer - but if Oracle really wants to sell products on BTRFS, then Oracle should keep most of the developers in house, so Oracle can tailor BTRFS the way Oracle wants it.

              Again, I do not claim these are facts. I just guess.

              When I do not guess, and I claim facts, I always back things up with research papers, etc. But these thoughts of mine, are not facts. Just guesses. Guesses are not FUD. FUD is something like "I have heard somewhere from someone that Oracle is killing off BTRFS, you better migrate now to ZFS" - that is a lie and FUD. I have not read from an Oracle executive that BTRFS is getting killed. It would be a lie and FUD.

              Comment


              • Originally posted by kraftman View Post
                You must show them to proof such papers exists, otherwise it's nothing more then FUDing and trolling. However, SUN FUDs.
                No I must not. FUD article on wikipedia does not specify which research journals the papers must be from. You are wrong on this.


                Originally posted by kraftman View Post
                I don't understand. While Solaris is dead then it can't be better then Linux which is not dead. Do you mean something which is dead is better then something which is not?
                Again (for the fourth(?) time): please specify in what way Solaris is dead? Do you mean it is not sold anymore? I do not know in which way you mean Solaris is dead. How can I show you are wrong, when you dont specify anything? It is like:
                -Kraftman, I think of something (I will not tell you what I think of) but I want you to disprove me. Go ahead.

                Do you want proof that Solaris is still sold? What do you want me to show? Can you please specify? Do you claim that Solaris is not sold anymore? What is it that you claim?

                Comment


                • Jalyst wrote
                  "Since when is solaris dead?
                  Still looks pretty healthy to me. "

                  Originally posted by kraftman View Post
                  According to some link it's dead already. However, read some previous posts.
                  Kraftman, you can not just post to any link. That is not proof at all. In that case, you can start a blog of your own, and write what you want, and then post to your blog. That is wrong to do (circular argument).

                  Instead, you must find credible links. Most credible are researchers and professors. If lots of professors and researchers say something, then it is more credible than if you post to an anonymous blog post. Also, other good and famous experts such as developers are credible.

                  If I want to post something negative about Linux, then I can not post to a Linux competitor such as a Sun blog. That would not be credible. Instead I must post to Linux people saying bad things about Linux - THAT is credible.

                  If you want to post something negative about Solaris, then you must find credible links. Such as; Solaris guys saying something bad about Solaris. Not Linux people saying bad things about Solaris - that is not credible. Or, if you can find some Linux developers looking at the Solaris code and they say the Solaris source code is bad. In short, try to find famous experts or researchers that say something.

                  You know, it is unlikely that a researcher say lies and false things, because other professors look at a research paper and if the professors find lies, the research paper will not be accepted into the scientific journal. Therefore, it is unlikely that research papers posts lies.

                  Also, other good sources are official benchmarks, because you reproduce the benchmarks. It is not possible to lie with benchmarks.

                  In short, you must find links that are probably not lies. Researchers (probably) dont lie. Official white papers dont lie. etc. But, competitors are not credible at all - they have every reason to lie. So your post about "Solaris is dead" is not credible. If you have Solaris developers saying that Solaris is dead, then it is a more credible link and more preferable.

                  Jesus, Kraftman. Don't you know this? Have you ever studied at university? Must I tell you everything about how you do higher University studies? Havent you studied "scientific theory" and how to do research and establish knowledge?

                  Comment


                  • "OpenSolaris is dead???? No!"
                    Originally posted by kraftman View Post
                    Yeah, right! When will be next release?
                    As I have written here before: OpenSolaris source code is alive and well. It is the basis for next gen Solaris 11. Solaris 11 is on it's way and will maybe be announced next month on Oracle Open World.

                    OpenSolaris distro (which is based on parts of the OpenSolaris source code) is another thing. Maybe Oracle has killed the OpenSolaris distro. We dont know. Maybe Oracle will tell us the status of OpenSolaris distro, next month?

                    In any case, there are several OpenSolaris community distros out there: Schillix, Nexenta, Korona, Milax, etc.

                    Comment


                    • Originally posted by jalyst View Post
                      Stupid editing time-limit...
                      Yes I do, in what context?, yes it is.
                      Because you have not read this thread from the beginning, I post this article again. Here is a link to a research paper about how XFS, JFS, ReiserFS, etc does not protect your data - which is what I have claimed earler.
                      http://www.zdnet.com/blog/storage/ho...ta-at-risk/169

                      I have also posted several research papers which backs up my claim about ZFS giving good protection, whereas common filesystems (including Raid-5 and raid-6) do not offer good data protection.

                      Do you want me to repost those research papers for you?


                      Nevermind Kraftman, claiming I do not post to research papers., He just lies about me

                      Comment


                      • Originally posted by kebabbert View Post
                        I am only saying that I believe if Oracle is serious about BTRFS, they should have dedicated far more resources internally than one full time paid developer. That is not FUD. It is true. And I do make it clear it is my BELIEF. I do not have proof or evidence that Oracle is abandoning BTRFS. It is just a guess from my side.
                        Your belief is wrong, Yan Zheng (Oracle), chris mason (Oracle) Josef Bacik (Red Hat)... seem full time dedicated. Look at the links, stop writing wrong data.


                        Your opinion is that "Oracle should have assigned a whole team of developers to BTRFS" but you affirmated this: "But guess what? There is only one single developer on BTRFS. Oracle is not interested in BTRFS, now that they got the best: ZFS" which is totally FUD, you don't know how developing on btrfs is going on.

                        On the other hand, linux is developed by many companies, why oracle has to put all the effort on btrfs. They simply continued the project of btrfs. A lot of companies will develop and will benefit.

                        Originally posted by kebabbert View Post
                        And I also say that normally, a company does not develop two identical products. And many people agrees that ZFS is superior to BTRFS. So I am asking, why develop an inferior copy, instead of betting on the original ZFS?
                        Easy, I have already posted it, there are linux customers that want btrfs.

                        Originally posted by kebabbert View Post
                        I make it clear that I guess and I think so. I do not claim to have it as facts. I know there are more than one full time paid developer - but if Oracle really wants to sell products on BTRFS, then Oracle should keep most of the developers in house, so Oracle can tailor BTRFS the way Oracle wants it.

                        Again, I do not claim these are facts. I just guess.
                        You affirmed things that now you are claiming that were just guesses, like:

                        -There is only one single developer on BTRFS. Oracle is not interested in BTRFS, now that they got the best: ZFS

                        -Linux is really bad as a Large Enterprise server. It is not because of the bad filesystems, but because of limitations in the Linux kernel

                        Now you are beginning t to change, good! i am glad. But please don't say that you were guessing because clearly you were asserting.

                        Comment


                        • Originally posted by kraftman View Post
                          http://osdir.com/ml/os.solaris.opens.../msg00020.html

                          Solaris feels so bloated and forking processes on it is so slow. Also,some applications consume a lot more CPU time on Solaris.Why is that, Korvar?
                          Now, this is a good link, Kraftman. This is credible because Solaris developers are saying things about Solaris. I do accept this link from you. I do not reject this link from you. (You always reject all my links).

                          According to this link, one guy can see process forking on Solaris just by looking at the output on the command line. That clearly shows some problem. The Solaris developer tries to debug and find the issue, but there is never any conclusion. The solaris developer just say "we have not focused on doing small things, maybe Linux is faster on this" and then tries to debug the problem. But the guy who reported the bug stops posting in the list. So we dont know what the problem was.

                          But I have never claimed that Solaris is faster than Linux on everything. Actually, I wrote that Linux may be faster on small workloads, but when we increase the workload to many cores, then Linux suffers and I gave proof to SAP official benchmarks.

                          But that was a good link, Kraftman. Continue like that. Find links from Solaris developers, and I will most probably accept your links.

                          Comment


                          • Originally posted by kebabbert View Post
                            But I have never claimed that Solaris is faster than Linux on everything. Actually, I wrote that Linux may be faster on small workloads, but when we increase the workload to many cores, then Linux suffers and I gave proof to SAP official benchmarks.
                            Those SAP official benchmarks didn't proof a thing. It was two unrelated benchmarks done on different hardware with different database used. Neither proof nor paper describing Linux scale worse then Solaris on such workload.

                            But that was a good link, Kraftman. Continue like that. Find links from Solaris developers, and I will most probably accept your links.
                            Solaris developers lie and FUD, so I don't believe they will be fair with others and I don't believe they'll saying true things about Solaris. I'm also not interested in proving you anything. Like I said, I can post a link sometimes etc.

                            Comment


                            • Originally posted by kebabbert View Post
                              No I must not. FUD article on wikipedia does not specify which research journals the papers must be from. You are wrong on this.
                              It seems papers backing up your claims doesn't exist. At least, I couldn't find them. Now, could you show them to backup your claims?

                              Again (for the fourth(?) time): please specify in what way Solaris is dead? Do you mean it is not sold anymore? I do not know in which way you mean Solaris is dead. How can I show you are wrong, when you dont specify anything? It is like:
                              -Kraftman, I think of something (I will not tell you what I think of) but I want you to disprove me. Go ahead.
                              If I'm correct, I don't have to specify you a thing. In this link I gave someone claims Solaris is dead. I'm just pasting, why are you asking me?

                              Kraftman, you can not just post to any link. That is not proof at all. In that case, you can start a blog of your own, and write what you want, and then post to your blog. That is wrong to do (circular argument).
                              I'm aware of this. However, how is this different of showing some Solaris guys or fanboys articles? Or public relations talk? Those are usually even worse then some blog post. Afaik you even gave a link to Bonwick's blog and he FUDs.

                              Comment


                              • Originally posted by Jimbo View Post
                                Your belief is wrong, Yan Zheng (Oracle), chris mason (Oracle) Josef Bacik (Red Hat)... seem full time dedicated. Look at the links, stop writing wrong data.
                                Of course I will stop writing wrong things, if I am not correct. I dont want to lie. You have posted some links, could you please point out where it says that those developers are full time paid by Oracle? Yes, there are developers active with BTRFS - I explained that earlier.

                                But I read in a link I got here, where Chris Mason said in an interview, something like: he "was the only full time paid developer at Oracle". I do not lie about this. He also said there are part time developers on BTRFS. But he explicitly said he was the only full time paid developer. I do not make this up, nor lie about this.

                                If you can show that there are several full time paid developers at Oracle, or an Oracle team, I will stop say that there is only one full time Oracle developer. Of course. I am not a liar or FUDer.


                                Your opinion is that "Oracle should have assigned a whole team of developers to BTRFS" but you affirmated this: "But guess what? There is only one single developer on BTRFS. Oracle is not interested in BTRFS, now that they got the best: ZFS" which is totally FUD, you don't know how developing on btrfs is going on.
                                I read in an interview from Chris Mason where he said he is the only full time paid developer - this is not FUD nor lie.

                                Oracle is not interested in BTRFS: there are several people thinks like this, not only me. At least one analyst said that. I write what I read: "from a business perspective, it is bad to sell two identical products" - said the analyst. This is also not FUD.


                                Easy, I have already posted it, there are linux customers that want btrfs.
                                I agree on this. But it would be easy for Oracle to say "ZFS is ready now, buy it now"


                                You affirmed things that now you are claiming that were just guesses, like:

                                -There is only one single developer on BTRFS.
                                This is not a guess. I read this, from the BTRFS main architect.


                                Oracle is not interested in BTRFS, now that they got the best: ZFS
                                This is a guess. I have not read any Oracle executive saying this. But I read an analyst saying this. But the analyst does not know, he is also guessing. So, yes, this is a guess. When I guess, I always write something like "in my opinion" or "in my point of view" or "I believe" etc


                                -Linux is really bad as a Large Enterprise server. It is not because of the bad filesystems, but because of limitations in the Linux kernel
                                This is not a guess. A Senior Storage expert says this. As other experts also says.


                                Now you are beginning t to change, good! i am glad. But please don't say that you were guessing because clearly you were asserting
                                Some things I was guessing. Other things I was asserting. And I do change about "only one full time devleoper at Oracle" if you can prove me wrong. The other things I have not changed. If you prove me wrong on other things, then I stop assert those things too, of course.

                                If you can prove me wrong on anything, I stop say so. At once. For instance, I do not say anymore that Linux has the graphics in the kernel, because someone posted links that disproved that. (But I read an Linux article where they said that Linux will have the graphic inside the Kernel, so I dont make that up or lie. I read that). If I see someone say that Linux kernel has graphics, then I correct them now: "No, Linux only has some parts in the kernel."

                                Comment

                                Working...
                                X