CVS redhat memory problem - Redhat

This is a discussion on CVS redhat memory problem - Redhat ; Hi, We have a very powerful server at work running only CVS and a licensing server. It's been a month now that after 3-4 days of uptime it really slows down and the CVS users complain that it's really slow ...

+ Reply to Thread
Results 1 to 2 of 2

Thread: CVS redhat memory problem

  1. CVS redhat memory problem

    Hi,

    We have a very powerful server at work running only CVS and a licensing
    server. It's been a month now that after 3-4 days of uptime it really slows
    down and the CVS users complain that it's really slow connecting to it and
    retrieving files from it. Here is a snapshot of the memory

    72 processes: 70 sleeping, 1 running, 0 zombie, 1 stopped
    CPU0 states: 0.0% user, 0.0% system, 0.0% nice, 100.0% idle
    CPU1 states: 0.0% user, 0.4% system, 0.0% nice, 99.1% idle
    Mem: 513980K av, 487936K used, 26044K free, 0K shrd, 80492K
    buff
    Swap: 2096472K av, 12764K used, 2083708K free 278640K
    cached


    Also, remote SSH fails to this box after 3-4 days of uptime too. Any
    suggestions? Is there any way to free the cached memory pages (Force it)?
    Because running the command top doesn't show any processes using up that
    much memory. Is having 26044K (out of 512 Megs) normal?

    Thank you
    Maziar A.



  2. Re: CVS redhat memory problem

    "Maziar Aflatoun" writes:
    >Hi,
    >
    >We have a very powerful server at work running only CVS and a licensing
    >server. It's been a month now that after 3-4 days of uptime it really slows
    >down and the CVS users complain that it's really slow connecting to it and
    >retrieving files from it. Here is a snapshot of the memory
    >
    >72 processes: 70 sleeping, 1 running, 0 zombie, 1 stopped
    >CPU0 states: 0.0% user, 0.0% system, 0.0% nice, 100.0% idle
    >CPU1 states: 0.0% user, 0.4% system, 0.0% nice, 99.1% idle
    >Mem: 513980K av, 487936K used, 26044K free, 0K shrd, 80492K
    >buff
    >Swap: 2096472K av, 12764K used, 2083708K free 278640K
    >cached
    >
    >
    >Also, remote SSH fails to this box after 3-4 days of uptime too. Any


    Fails how? We can't read your mind. Do you get an error message? Does
    it just hang?

    >suggestions? Is there any way to free the cached memory pages (Force it)?


    The memory status, above, is normal. The cached pages will be automatically
    claimed as necessary.

    >Because running the command top doesn't show any processes using up that
    >much memory. Is having 26044K (out of 512 Megs) normal?


    The file system will use any free memory to cache file contents to make
    normal file accesses more efficient. If needed for process pages, the
    VM system will reclaim cached file content (flushing as necessary).

    You don't say which redhat distribution you are running.

    I presume your CVS users are using server: mode to connect to the
    host. In this case, I would attach strace (using the -tt -f and -o flags)
    to xinetd, then create a client CVS connection to the host (checkout a
    single file, for example). You'll want to do this at a time when no other
    clients are using the host to avoid a messy strace.

    After the cvs client operation has completed, kill the strace and examine
    the system call trace to find out which operations take the most time. I
    suspect you'll find they are network related (e.g. reverse IP DNS lookups or
    something similar).

    scott
    >
    >Thank you
    >Maziar A.
    >
    >


+ Reply to Thread