Find the answer to your Linux question:
Results 1 to 10 of 10
would like to know how to make the average queue length of HD. and therefore that each process be using? anyone can do this by prompt?...
  1. #1
    Just Joined!
    Join Date
    May 2010
    Posts
    4

    queue HD

    would like to know how to make the average queue length of HD.

    and therefore that each process be using?

    anyone can do this by prompt?

  2. #2
    Linux Guru Rubberman's Avatar
    Join Date
    Apr 2009
    Location
    I can be found either 40 miles west of Chicago, or in a galaxy far, far away.
    Posts
    8,974
    As far as I know, you can only view what the average queue length is (see the iostat command). If it can be set, then you probably need to modify the kernel to do so - not recommended. Why do you want to do this? Is there a problem that you are thinking this will resolve?
    Sometimes, real fast is almost as good as real time.
    Just remember, Semper Gumbi - always be flexible!

  3. #3
    Just Joined!
    Join Date
    May 2010
    Posts
    4
    Quote Originally Posted by Rubberman View Post
    As far as I know, you can only view what the average queue length is (see the iostat command). If it can be set, then you probably need to modify the kernel to do so - not recommended. Why do you want to do this? Is there a problem that you are thinking this will resolve?
    I have a software running on a server, a client, but this server runs over an Outsourced software, and a few more things from the client.

    In short, the server is with some peaks to catch, and do not know what yet, give one accuses the other of guilt. I'm sure that's not our software, but I have to prove it.

  4. #4
    Linux Guru Rubberman's Avatar
    Join Date
    Apr 2009
    Location
    I can be found either 40 miles west of Chicago, or in a galaxy far, far away.
    Posts
    8,974
    So, there is a latency problem, or something else? Disc queue lengths can contribute to that if the OS is taking too much time servicing them to the detriment of applications that also need I/O. You can run iostat with options that will run continuously at specified intervals in order to get some view of how it is behaving over time. There are also commercial network monitoring tools you can get which help identify causes of performance problems, such as HP's Glance, which I have used to great effect in the past when I was doing performance analysis of large scale distributed systems. It isn't cheap, however... and I think it is only available for HP-UX operating systems. I'll post again after I look for some other tools for you. CA Unicenter is another option which I have also used in the past.
    Sometimes, real fast is almost as good as real time.
    Just remember, Semper Gumbi - always be flexible!

  5. #5
    Just Joined!
    Join Date
    May 2010
    Posts
    4
    Thank you for your attention. I await your return.

    I'm sorry, maybe my English is not good, because I am Brazilian and can not speak English very well.

  6. #6
    Linux Guru Rubberman's Avatar
    Join Date
    Apr 2009
    Location
    I can be found either 40 miles west of Chicago, or in a galaxy far, far away.
    Posts
    8,974
    Well, all I can about CA's current web site is BLARGH! What dren! Absolutely no useful information to be found there. Why anyone would want to buy anything of theirs is beyond me. Apparently Unicenter (purchased from DEC before HP bought them) has been merged into something else. It sounds like you have to purchase the entire enterprise management suite in order to get the performance analysis tools of Unicenter.

    HP at least has some useful online information about Glance and other tools, but also unforunately it seems Glance isn't available for Linux systems, as far as I can tell. Too bad. It is a great tool.

    So, here is one Linux link that may be helpful: SYSSTAT
    I'll post some more if I find them. Linux has a lot of little tools that give a lot of information about system performance and resource utilization; however, there is a dearth of really high-class tools to help visualize and drill down into problem areas in a way that is useful for root-cause analysis of problems such as yours.
    Sometimes, real fast is almost as good as real time.
    Just remember, Semper Gumbi - always be flexible!

  7. #7
    Linux Guru Rubberman's Avatar
    Join Date
    Apr 2009
    Location
    I can be found either 40 miles west of Chicago, or in a galaxy far, far away.
    Posts
    8,974
    Quote Originally Posted by malhuco View Post
    Thank you for your attention. I await your return.

    I'm sorry, maybe my English is not good, because I am Brazilian and can not speak English very well.
    My second language is Spanish, which I am fluent in. I can usually understand Portuguese, but I don't speak it. In any case, these forums are an English-only venue, so just do the best you can.
    Sometimes, real fast is almost as good as real time.
    Just remember, Semper Gumbi - always be flexible!

  8. #8
    Just Joined!
    Join Date
    May 2010
    Posts
    4

    Talking

    Well, I am now installing the Sysstat, my sincere thanks for your attention and dedication to help.

  9. #9
    Linux Enthusiast Mudgen's Avatar
    Join Date
    Feb 2007
    Location
    Virginia
    Posts
    623
    Quote Originally Posted by Rubberman View Post
    HP at least has some useful online information about Glance and other tools, but also unforunately it seems Glance isn't available for Linux systems, as far as I can tell. Too bad. It is a great tool.
    GlancePlus is part of HP OpenView, and is available for Linux platforms. Trial download at
    https://h10078.www1.hp.com/cda/hpms/display/main/hpms_content.jsp?zn=bto&cp=1-11-15-28^9637_4000_100

    You need an HP Passport ID to download. Free registration available via the download page.

  10. #10
    Just Joined!
    Join Date
    May 2005
    Posts
    8

    Arrow

    consider setting up munin to measure your hosts' performance, its very good

    you do need to put the central munin system on a computer of its own if you are measuring more than say 10 servers as it can be a cpu and disk hog

    we use cacti for networks, munin for servers and nagios for alerting/monitoring

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  
...