Linux.com

coop

coop

  • Moderator
  • Posts: 87
  • Member Since: 19 Nov 09
  • Last Logged In: 21 hours ago

Latest Posts

Posted by
Topic
Post Preview
Posted
  • coop
    RE: Will LFS201 materials be updated to match the new LFCS Domains and Competencies V2.16 (Feb 2, 2016)
    Actually the person complained about things like cutting and pasting text, which is not considered advanced material and is definitely not in the course. Nothing about advanced material has every been specifically mentioned as being on the test, only people are worried about it being on the test. And we keep reminding people they get a free retake so they can get a pretty good look at type of material. As to " Is it not common sense to revamp study material first and then publish revised scope of exam? It is simple, really." .... The answer is no, this is not the way things can work. The group making exams is independent of the group designing training materials. It is as if they were two separate organizations. As a result, the training people (who are the people on this forum such as myself) have to react to *published* lists of coverage areas. In principle you can take training from any vendor who wants to provide it, just as you can for redhat certification etc. competition is allowed. Unfortunately, the revised list of domains is too repetitious between LFCE and LFCS as has been commented on in this and other threads multiple times, and you have to rely on the published weights between the exams. I wish it were otherwise and I have complained about it, but until they come out with another revision nothing will change :(
    Link to this post 21 hours ago

    Actually the person complained about things like cutting and pasting text, which is not considered advanced material and is definitely not in the course. Nothing about advanced material has every been specifically mentioned as being on the test, only people are worried about it being on the test. And we keep reminding people they get a free retake so they can get a pretty good look at type of material.

    As to " Is it not common sense to revamp study material first and then publish revised scope of exam? It is simple, really." ....

    The answer is no, this is not the way things can work. The group making exams is independent of the group designing training materials. It is as if they were two separate organizations. As a result, the training people (who are the people on this forum such as myself) have to react to *published* lists of coverage areas. In principle you can take training from any vendor who wants to provide it, just as you can for redhat certification etc. competition is allowed.

    Unfortunately, the revised list of domains is too repetitious between LFCE and LFCS as has been commented on in this and other threads multiple times, and you have to rely on the published weights between the exams. I wish it were otherwise and I have complained about it, but until they come out with another revision nothing will change :(

  • coop
    RE: Lab 26.1 - Any tips for doing the lab? I don't see any messages of OOM-killer in my logs :-(
    "If I choose to stress requesting more memory than is available " is where you probably went wrong. You probably had page cache memory which could be freed without causing the system to truly run out. If you ask for very large amounts the system is smart enough to give up. What you need to do is if you have say 4 GB of mem, ask for 4 GB (or 3.8 GB or so) as that can not be satisfied and keep everything else running. if you were to ask for 20 GB it would probably just tell you nope. To see the oom messages you can do "dmesg -w" (or look at system logs)
    Link to this post 1 day ago

    "If I choose to stress requesting more memory than is available " is where you probably went wrong. You probably had page cache memory which could be freed without causing the system to truly run out.

    If you ask for very large amounts the system is smart enough to give up. What you need to do is if you have say 4 GB of mem, ask for 4 GB (or 3.8 GB or so) as that can not be satisfied and keep everything else running. if you were to ask for 20 GB it would probably just tell you nope.

    To see the oom messages you can do "dmesg -w" (or look at system logs)

  • coop
    RE: Lab 21.1 - Can someone explain signals and signal handlers please?
    Please note we don't expect students to be C programmers, although it is nice if they are. The only reason the source code is included in the text (instead of just a downloadable file) is to avoid the problems some students have when downloading on windows machines; instead they can cut and paste. When a signal is blocked it is *temporarily* prevented from being handled; as soon as the blockage is removed the "handler" deals with it, doing whatever should be done, including possibly killing the program. Think of it as a "do not disturb" sign on the door; when the sign is removed the hotel room can be cleaned. The signal should never be "dropped"; it should be dealt with whenever the blockage is removed. This is a lab where we are trying to teach something, not just quiz the knowledge of what is in the section earlier.
    Link to this post 4 days ago

    Please note we don't expect students to be C programmers, although it is nice if they are. The only reason the source code is included in the text (instead of just a downloadable file) is to avoid the problems some students have when downloading on windows machines; instead they can cut and paste.

    When a signal is blocked it is *temporarily* prevented from being handled; as soon as the blockage is removed the "handler" deals with it, doing whatever should be done, including possibly killing the program. Think of it as a "do not disturb" sign on the door; when the sign is removed the hotel room can be cleaned. The signal should never be "dropped"; it should be dealt with whenever the blockage is removed.

    This is a lab where we are trying to teach something, not just quiz the knowledge of what is in the section earlier.

  • coop
    RE: Will LFS201 materials be updated to match the new LFCS Domains and Competencies V2.16 (Feb 2, 2016)
    I just want to remind people that this class (LFS201) was never claimed to have *everything* you need. In particular you should have taken LFS101 (for free at edx) or equivalent which is where editors, text manipulation, getting help and many other topics are covered as is explained in the beginning of this course and in the certification prep guides. If advanced topics are not covered that is a different question.
    Link to this post 5 days ago

    I just want to remind people that this class (LFS201) was never claimed to have *everything* you need. In particular you should have taken LFS101 (for free at edx) or equivalent which is where editors, text manipulation, getting help and many other topics are covered as is explained in the beginning of this course and in the certification prep guides. If advanced topics are not covered that is a different question.

  • coop
    RE: Lab 4.1 fake_service.txt
    This was corrected (at least in the English version) last summer, and in this very thread there is a link to a web site (at training.linuxfoundation.org) where you can download a perfect copy without the dos2unix conversion. Furthermore, when I download the script from the download button it is also fine. If it is not, it is because you are downloading under Windows and it is doing a unix->dos conversion on you, which we can't do anything about. Furthermore, in the description of the exercise the actual full text of the script is there, and you can simply cut and paste it (in any language). Unfortunately, when you use Windows programs (such as some browser downloads and pdf viewers such as acroread) these kinds of problems can occur as they try to "improve" upon unix customs. As far as "stick notes" go, we update the course frequently to fix minor problems such as these and as long as we don't introduce new material or make more substantive changes you won't even notice.
    Link to this post 02 Feb

    This was corrected (at least in the English version) last summer, and in this very thread there is a link to a web site (at training.linuxfoundation.org) where you can download a perfect copy without the dos2unix conversion. Furthermore, when I download the script from the download button it is also fine. If it is not, it is because you are downloading under Windows and it is doing a unix->dos conversion on you, which we can't do anything about. Furthermore, in the description of the exercise the actual full text of the script is there, and you can simply cut and paste it (in any language).

    Unfortunately, when you use Windows programs (such as some browser downloads and pdf viewers such as acroread) these kinds of problems can occur as they try to "improve" upon unix customs.

    As far as "stick notes" go, we update the course frequently to fix minor problems such as these and as long as we don't introduce new material or make more substantive changes you won't even notice.

  • coop
    RE: Will LFS201 materials be updated to match the new LFCS Domains and Competencies V2.16 (Feb 2, 2016)
    Thanks for giving the reference, let me just amplify a bit and encourage people to follow the link he/she gave (repeated below). This question has come up before; unfortunately the forum format does not make searching previous threads easy, and we are going to have an upgrade in the near future to make this more productive. In the thread: https://www.linux.com/community/forums/lfs201-class-forum/is-this-course-deep-enough-to-pass-the-exam-maybe-not-for-the-new-exam-216 this came up and we explained that the exam staff did a less than ideal step of making the areas covered by LFCS and LFCE very similar but altering the *weights* assigned. This is confusing and makes it more difficult to sanely prepare than it should be. however, using some common sense about what is reasonable to expect on an LFCS exam can help a lot (it is performance based, no multiple choice and things have to be doable in a short period of time.) Furthermore, it is not a trivia test. Finally you can do a retake for free, so you can try and limit your anxiety on the first go around until you see what it is really liike.
    Link to this post 02 Feb

    Thanks for giving the reference, let me just amplify a bit and encourage people to follow the link he/she gave (repeated below).

    This question has come up before; unfortunately the forum format does not make searching previous threads easy, and we are going to have an upgrade in the near future to make this more productive.

    In the thread:

    https://www.linux.com/community/forums/lfs201-class-forum/is-this-course-deep-enough-to-pass-the-exam-maybe-not-for-the-new-exam-216

    this came up and we explained that the exam staff did a less than ideal step of making the areas covered by LFCS and LFCE very similar but altering the *weights* assigned. This is confusing and makes it more difficult to sanely prepare than it should be. however, using some common sense about what is reasonable to expect on an LFCS exam can help a lot (it is performance based, no multiple choice and things have to be doable in a short period of time.) Furthermore, it is not a trivia test. Finally you can do a retake for free, so you can try and limit your anxiety on the first go around until you see what it is really liike.

  • coop
    RE: When is the course going to be updated again?
    Having exam and course staffs separated is a good standard practice; it is also done by some other organizations that provide both, such as Red Hat. In some circumstances and jurisdictions, it is even a legal requirement! Otherwise there is a tendency to only "teach the test" which is counterproductive. Furthermore, it allows third party organizations and companies to sell training on an equal footing. The Linux Foundation is not interested in having a monopoly on training for exam preparation, we are not that kind of entity.
    Link to this post 01 Feb

    Having exam and course staffs separated is a good standard practice; it is also done by some other organizations that provide both, such as Red Hat. In some circumstances and jurisdictions, it is even a legal requirement!

    Otherwise there is a tendency to only "teach the test" which is counterproductive.

    Furthermore, it allows third party organizations and companies to sell training on an equal footing. The Linux Foundation is not interested in having a monopoly on training for exam preparation, we are not that kind of entity.

  • coop
    RE: Lab 11.1 - my /dev/shm is not tmpfs [SOLVED - it`s normal]
    This is normal. Exactly what is on tmpfs filesystems depends on your distribution; you did not mention what that is. For example on my (RHEL 7) system: $ df -h | grep tmpfs devtmpfs devtmpfs 7.8G 0 7.8G 0% /dev tmpfs tmpfs 7.8G 152M 7.7G 2% /dev/shm tmpfs tmpfs 7.8G 9.1M 7.8G 1% /run tmpfs tmpfs 7.8G 0 7.8G 0% /sys/fs/cgroup tmpfs tmpfs 1.6G 40K 1.6G 1% /run/user/1000 You can have multiple tmpfs ramdisks mounted in fact as shown above. I've yet to see a distribution that doesn't use /dev/shm, but it's not essential to have it. Also note they are really not 7.8 GB in size here (which is half of my RAM). they are allowed to grow to that big but the size is dynamical.
    Link to this post 31 Jan

    This is normal. Exactly what is on tmpfs filesystems depends on your distribution; you did not mention what that is. For example on my (RHEL 7) system:

    $ df -h | grep tmpfs
    devtmpfs devtmpfs 7.8G 0 7.8G 0% /dev
    tmpfs tmpfs 7.8G 152M 7.7G 2% /dev/shm
    tmpfs tmpfs 7.8G 9.1M 7.8G 1% /run
    tmpfs tmpfs 7.8G 0 7.8G 0% /sys/fs/cgroup
    tmpfs tmpfs 1.6G 40K 1.6G 1% /run/user/1000

    You can have multiple tmpfs ramdisks mounted in fact as shown above.

    I've yet to see a distribution that doesn't use /dev/shm, but it's not essential to have it. Also note they are really not 7.8 GB in size here (which is half of my RAM). they are allowed to grow to that big but the size is dynamical.

  • coop
    RE: When is the course going to be updated again?
    Just to be clear, Flavia and I are part of the team responsible for the course. We are not the exam team; there is a firewall between. The course has already been updated for the new exam versions, according to the exam people's public information. This class does receive periodic updates, mostly bug fixes and changes to labs and exposition that clarify. We do that every several months; doing it more often just confuses people as they may lose bookmarks etc, and we announce when new updates are done. More major updates, such as completely new content are less often and are strongly announced.
    Link to this post 31 Jan

    Just to be clear, Flavia and I are part of the team responsible for the course.

    We are not the exam team; there is a firewall between.

    The course has already been updated for the new exam versions, according to the exam people's public information.

    This class does receive periodic updates, mostly bug fixes and changes to labs and exposition that clarify. We do that every several months; doing it more often just confuses people as they may lose bookmarks etc, and we announce when new updates are done. More major updates, such as completely new content are less often and are strongly announced.

  • coop
    RE: Knowledge check 8.1
    Thanks Istimsak for shedding some light. Let me explain further. If you look at your disk when the system is not running, /dev is an empty directory. It is an in-memory pseudofilesystem (a ram fs) mounted at boot and populated by udev as devices are found. (read the udev chapter). The entries in /dev are indeed *not* normal files; they are character nodes; do ls -l and you will see block devices have a "b" in the first character (like disk partitions such as /dev/sda1) and character devices have a "c" (like /dev/lp0 for first printer). Network devices do not have device nodes. Many device nodes do not correspond to hardware but are "software devices", an abstraction useful to the kernel, such as /dev/null or /dev/fuse. You have two character devices under /dev/ati; /dev/ati/card0 and /dev/ati/card1. Some video cards create character devices, some don't. They would be used for streaming bytes to the card. I'm not familiar with the ati cards and it will also depend on whether you are using an in-kernel device driver or a proprietary closed source one from ati. On the two systems I am looking at right now, both have Intel graphics chips on board and neither seems to have a character device associated with them. By the way, unless you know what you are doing, avoid doing any I/O directly to anything under /dev; you are quite likely to do something like corrupt your hard disk or fry something else. For example, when you read/write a file the filesystem will eventually do I/O with /dev/sda1 for example, but if you do it directly you will trash the filesystem.
    Link to this post 31 Jan

    Thanks Istimsak for shedding some light. Let me explain further.

    If you look at your disk when the system is not running, /dev is an empty directory. It is an in-memory pseudofilesystem (a ram fs) mounted at boot and populated by udev as devices are found. (read the udev chapter). The entries in /dev are indeed *not* normal files; they are character nodes; do ls -l and you will see block devices have a "b" in the first
    character (like disk partitions such as /dev/sda1) and character devices have a "c" (like /dev/lp0 for first printer).

    Network devices do not have device nodes.

    Many device nodes do not correspond to hardware but are "software devices", an abstraction
    useful to the kernel, such as /dev/null or /dev/fuse.

    You have two character devices under /dev/ati; /dev/ati/card0 and /dev/ati/card1. Some video cards create character devices, some don't. They would be used for streaming bytes to the card. I'm not familiar with the ati cards and it will also depend on whether you are using an in-kernel device driver or a proprietary closed source one from ati. On the two systems I am looking at right now, both have Intel graphics chips on board and neither seems to have a character device associated with them.

    By the way, unless you know what you are doing, avoid doing any I/O directly to anything
    under /dev; you are quite likely to do something like corrupt your hard disk or fry something else. For example, when you read/write a file the filesystem will eventually do I/O with /dev/sda1 for example, but if you do it directly you will trash the filesystem.

Who we are ?

The Linux Foundation is a non-profit consortium dedicated to the growth of Linux.

More About the foundation...

Frequent Questions

Linux Training / Board

/** BC-056 Ameex changes to add tracking code - 2016-01-22 **/ ?>