发布新日志

  • Windows性能计数器

    2010-02-26 14:55:55

    对象

    计数器

    分析

    processor

    %processor time

    建议阈值85%

    memory Available bytes

     

    建议阈值少于4MB需要添加内存;另外,又建议至少要有10%的物理内存值

    Pages reads/sec

     

    是指为解析硬页错误而读取磁盘的次数,如果该值一直持续较大,表明可能内存不足建议阈值305?),大数值表示磁盘读而不是缓存读

    Pages writes/sec

     

    是指为了释放物理内存空间而将页写入磁盘的次数

    Pages Input/sec

     

    指为解决页错误从磁盘上读取的页数

    Pages Output/sec

     

    是指为了释放物理内存空间而写入磁盘的页数;如果该值远远大于Pages Input/sec,可能有内存泄露

    Pages/sec

     

    是指为解析硬页错误从磁盘读取或写入磁盘的页数;建议阈值20

    Network interface

    Bytes received/sec

    (对于TCP/IP 该数据结合Bytes total/sec

     

    Bytes sent/sec

    该数据结合Bytes total/sec

     

    Bytes total/sec

    推荐不要超过带宽的50%

     

    Packets/sec

    根据实际数据量大小,无建议阈值,该数据结合Bytes total/sec

     

    Physical disk Disk reads/sec

    取决于硬盘制造商的规格,检查磁盘的指定传送速度,以验证此速度没有超出规格

     

    Disk writes/sec

    取决于硬盘制造商的规格,检查磁盘的指定传送速度,以验证此速度没有超出规格;又:上两值相加,应小于磁盘设备的最大容量

     

    %Disk Time

    建议阈值90%

     

    Current disk queue length

     

     

    Avg. disk queue length

    (如果使用RAID设备,%Disk Time计数器显示的值可以大于100%。如果大于100%,则使用Avg. disk queue length计数器决定正在等待磁盘访问的系统请求的平均数) 不超过磁盘数的1.5~2倍;如果上两值始终较高,可以考虑升级磁盘驱动器或将某些文件移动到其他磁盘或服务器

  • HP RAMS 介绍

    2008-08-05 17:36:50

    The Route Analytics Management System (RAMS) is an IP Route Analytics
    tool that listens to routing protocols and builds a real-time routing topology
    map. This map helps you visualize and understand the dynamic operation of
    the network.
    RAMS offers the following powerful contributions to network planning and
    analysis:
    • Unified, real-time routing topology view. View complex topologies
    hierarchically or by protocol, autonomous system (AS), or IGP area. The
    History Navigator window lets you play back a history of your routing
    topology changes.
    • Monitoring and alerts. Monitor vital service parameters (network
    churn, prefix flaps, and so on), watch for changes in specific end-to-end
    service paths and prefixes, and look for degrading redundancy. RAMS can
    also raise alerts on all watched parameters to head off costly outages.
    • Interactive routing analysis. Perform “before and after” comparisons
    and detailed event analysis using a comprehensive routing base and
    complete event history to help you rapidly establish the cause of the
    problem.
    • Planning support. Display network activity patterns to help you
    optimize performance and minimize unnecessary transit fees or bandwidth
    costs. You can simulate a link failure, or change link metric costs, to see
    how your routing topology will respond to specific failures or upgrades. You
    can also import and export these simulated changes, so you can manage
    multiple routing scenarios using external editors.
    • Reports. View trends and identify emerging issues before they become
    problems. You can generate web-based reports for any recorded time
    period, which show you key information about network health.

Open Toolbar