UNSOLVED

Enthusiast11

updated

16 years ago

1

7331

September 29th, 2010 08:00

Symmetrix performance

Hi All.

Can anyone please let me know how do we analyze the symmetrix perf using symstat output..also can anyone let me know how can we say by what values we say tha ports, devices etc are not overloaded. Also is there a way we can get the symm stat for a sepcific disk group, and find out what luns are having heavy ios and also please let me know how can we analyze the symstat output of cache...I really appreciate anyones help in this...

  • Boom1

    131 Posts

    1988

    1

    Posted September 29th, 2010 10:00

    I will suggest use SPA.

    yes, there is a option available for device group.

    Metric

    Description

    Host IOs/sec

    Number of host IO operations performed each second   by all Symmetrix devices, including writes and random and sequential reads.

    Host Reads/sec

    Number of host read operations performed each second   by all Symmetrix devices.

    Host Writes/sec

    Number of host write operations performed each second   by all Symmetrix devices.

    % Reads

    Percent of total read IO operations performed each   second by all of the Symmetrix devices.

    % Writes

    Percent of total write IO operations performed by   all of the Symmetrix devices.

    % Hit

    Percent of IO operations performed by all of the   Symmetrix devices, for which the read data was in cache and the write   operation could be sent directly to cache without having to wait for data   to be destaged from cache to the disks.

    Host MBs/sec

    The number of host MBs written and read by all of   the Symmetrix devices each second.

    Host MBs Read/sec

    The number of host MBs read by all of the Symmetrix   devices each second.

    Host MBs Written/sec

    The number of host MBs written by all of the Symmetrix   devices each second.

    FE Reqs/sec

    Data transfer between the director and the cache.   An IO may require multiple requests depending on IO size, alignment, or   both. The requests rate should be either equal to or greater than the   IO rate.

    FE Read Reqs/sec

    A read data transfer between the director and the   cache. An IO may require multiple requests depending on IO size, alignment,   or both. The requests rate should be either equal to or greater than the   IO rate.

    FE Write Reqs/sec

    A write data transfer between the director and the   cache. An IO may require multiple requests depending on IO size, alignment,   or both. The requests rate should be either equal to or greater than the   IO rate.

    BE IOs/sec

    Total IO from all BE directors to the disks per second.

    BE Reqs/sec

    A data transfer of a read or write between the cache   and the director.

    BE Read Reqs/sec

    A data transfer of a read between the cache and the   director.

    BE Write Reqs/sec

    A data transfer of a write between the cache and   the director.

    System WP Events/sec

    Number of times each second that write activity was   heavy enough to use up the system limit set for write tracks occupying   cache. When the limit is reached, writes are deferred until data in cache   is written to disk.

    Device WP Events/sec

    Number of times each second that the write-pending   limit for a specific Symmetrix device was reached.

    WP Count

    Number of system cache slots that are write pending.

    System Max WP limit

    The percent of the target % at which writes are delayed.   The range is from 40% to 80%.

    % Cache WP

    Percent of system cache that is write pending.

    Avg Fall Thru Time

    Average time it takes a cache slot in LRU0 to be   freed up. It is the average time from the first use of the contents to   its reuse by another address.

    FE Hit Req/sec

    Total requests from all FE directors per second that   were satisfied from cache.

    FE Read Hit Reqs/sec

    Total read requests from all FE directors per second   that were satisfied from cache.

    FE Write Hit Reqs/sec

    Total write requests from all FE directors per second   that were satisfied from cache.

    Prefetched Tracks/sec

    Tracks per second pre-fetched from disk to cache   upon detection of a sequential read stream.

    Destaged Tracks/sec

    Tracks per second saved into disks.

    FE Read Miss Reqs/sec

    Total read requests from all FE directors per second   that were misses. A miss occurs when the requested read data is not found   in cache.

    FE Write Miss Reqs/sec

    Total write requests from all FE directors per second   that were misses.  A miss occurs when the write   operation had to wait while data was destaged from cache to the disks.

  • Quincy561

    1281 Posts

    1988

    0

    Posted September 29th, 2010 12:00

    You can use symstat to monitor IO rates, but that won't reallty help you understand utilization unless you know what the limits are for the exact workload that is running.

    IO size and type matter a lot.  For example, an FA CPU could be maxed out at just 1,000 IOs/sec or the limit could be 20,000 depending on the IO size and type.

    The chart below shows a single IO type (read miss) maximum IOs/sec with various IO sizes.

    ScreenShot203.bmp

  • Booyah2

    2 Intern

    •

    185 Posts

    1988

    0

    Posted September 30th, 2010 06:00

    Good morning Enthusiasts1,

         There is a Forum document on this topic you may find very helpful. I have pasted it below, but you can also find it in the documents tab on this site.

    Performance Monitoring Using CLI Command SYMSTAT VERSION 7  Click to view document history

    Created on: Sep 2, 2010 7:04 AM by Thomas Moore

    - Last Modified:  Sep 2, 2010 11:31 AM by SymmetrixForumsModerator

    Product:

    All Symmetrix models, including Symmetrix VMAX

    Solutions Enabler


    Description:

    How to determine cache hit rates.

    How to determine the write pending count.

    How to monitor performance of your EMC Symmetrix.

    How to use the symstat command.

    Resolution:

    Refer to the EMC Solutions Enabler Symmetrix Array Management CLI Version 7.0 manual, P/N 300-002-939. The link is provided below. To find this manual, from the home page of Powerlink select Support > Technical Documentation and Advisories > Software ~ S ~ Documentation > Solutions Enabler > v7.0x.

    http://powerlink.emc.com/km/live1/en_US/Offering_Technical/Technical_Documentation/300-002-939_a06.pdf?mtcs=ZXZlbnRUeXBlPUttQ2xpY2tDb250ZW50RXZlbnQsZG9jdW1lbnRJZD0wOTAxNDA2NjgwNDBiODUwLGRvY3VtZW50VHlwZT1wZGYsbmF2ZU5vZGU9MGIwMTQwNjY4MDQwYjAxZV9Hcmlk

    Shown below is an example from the Solutions Enabler manual linked above, taken from page 174. "Statistics Overview" starts on page 171.

    Array statistics

    I/O and throughput statistics can be returned for all devices on a specific array by specifying the Symmetrix ID in the command-line, specifying the interval and the number of samples (counts) for the statistical run. The following example returns performance statistics for Symmetrix array 150:

    symstat -type REQUEST -sid 150 -i 10 -c 100

    The following statistics are returned:

    1. Host timestamp of the sample.

    2. Active Symmetrix device name and host physical device name (if the device is visible).

    3. Read cache request rate (requests per second).

    4. Write cache request rate (requests per second).

    5. Read throughput rate (kilobytes per second).

    6. Write throughput rate (kilobytes per second).

    7. Read/write cache hit ratio (percentage of write hits).

    8. Sequential read/write ratio (percentage of sequential reads).

    9. Write-pending track count.

  • Booyah2

    2 Intern

    •

    185 Posts

    1988

    1

    Posted September 30th, 2010 06:00

    Here's a really good performance class offered at EMC facility.

    Here's the training course catalogue:

    http://powerlink.emc.com/km/live1/en_US/Training/Supporting_Collateral/TRN_H1169_GED_paths.pdf?mtcs=ZXZlbnRUeXBlPUttQ2xpY2tDb250ZW50RXZlbnQsZG9jdW1lbnRJZD0wOTAxNDA2NjgwNTI4NTVmLGRvY3VtZW50VHlwZT1wZGYsbmF2ZU5vZGU9MGIwMTQwNjY4MDA3Y2IzN19Hcmlk

    Here's the course title below and outlined on pages 29 and 43 in the above catalogue..This is a 4 day class > very good class.

    Symmetrix Performance Workshop

    (Instructor-Led)

    Performance overview and measurements, tools, analysis, front-end, system/

    cache, back-end, business continuance

    Hope this helps,

    Thom

    If any of this information is helpful, please let us know by marking the replies appropriately.

  • 1988

    0

    Posted October 6th, 2010 10:00

    Thank you everyone for your valuble replies...

    Symstat output for the FAs

                      DIRECTOR  IO/sec     Cache             Requests/sec        % RW

    14:39:08                         Host     READ              WRITE       RW        Hits

    14:39:08       FA-4A          442         459                  119       579          92

    Symstat output for the Devices

    DEVICE                                  IO/sec             KB/sec            % Hits                   %Seq Num                   WP

    14:00:26                             READ WRITE    READ WRITE     RD WRT                    RD WRT                  Tracks

    14:00:39 DEV001 (19EF)         202      52      21040 18448        88 100                       13   35                     4636

    Can any one let me know how I can interpret the above values..and what are the things I would need to take care. Also how can I know if any of the values are peaked out.

    Thank you in advance...