Announcement Banner
UNSOLVED

AshJo

updated

14 years ago

A

AshJo

1 Rookie

•

60 Posts

0

1141

June 20th, 2012 08:00

SRDF/A

Hi,

When a RDF grp suspend,

What all parameters we have to look in, other than the link failure?

Why the device reach's max WP? How can I get into root cause of a dev reaching max WP and RDF grp getting suspended?

We have SPA, installed. What all the parameters which I can take into account to get the root cause. Say active cycle size, Uncommitted tracks...etc

  • anoopcr

    2 Intern

    •

    148 Posts

    535

    0

    Posted June 20th, 2012 09:00

    maximum cache limi that can be use for SRDF\A operation is 94 in DMX and

    in Vmax its 75...and also for each device, it has maximum limit of 5%

    cache utilization.If you dont have a much band width utilization for SRDF

    to catch up the writes to the R1 device,devices will reach its WP limit

    and that will cause srdf failure.

  • 535

    0

    Posted June 24th, 2012 01:00

    Ashwini,

    In Spa you can keep a track of wp limit and cache utilisation. Setup the email alerts for both.

    Regards,

    Saurabh

  • Jayd

    2 Posts

    535

    0

    Posted June 28th, 2012 14:00

    Our SRDF/A is also being suspended with alert Description: SRDF/A Session Dropped: System Max WP limit reached, and we're not allowed to throttle the host

    As per Anoop's comment does this mean there is not enough badwidth on the SRDF link causing write pending limits. On the FCIP tunnel we see only 2-3MB/s data crossing over on 2 tunnels whereas during  the adaptive copy session its up in 10-12 MB/s.

    So does not see like the bandwidth is concern here, its does seem to cache hitting the 75% limit.

    Any ideas?

  • PedalHarder

    4 Apprentice

    •

    465 Posts

    535

    0

    Posted June 28th, 2012 19:00

    There are no issues related to the message you received. It will be accurate. Do you mean it was only 2-3MB/sec when SRDF/A dropped compared to 10-12 when ADcopy was running?

    If you got a drop at such a low throughout rate, the first thing I would look at is the quality of the link. What is your packet loss ratio?

  • Jayd

    2 Posts

    535

    0

    Posted June 29th, 2012 10:00

    Yes, when SRDF/A is running I only see  traffic between 2-3 MB/s whereas when in ADCopy mode it jumps to 10-11 MB/s. I am trying to find where i can see the packet loss statistics. We are using Brocade 7800 FCIP router on both ends with compression on as below;

    -------------------------------------------------------------------------------

    Tunnel Circuit  OpStatus  Flags    Uptime  TxMBps  RxMBps ConnCnt CommRt  Met

    -------------------------------------------------------------------------------

    16     -         Up           c------       7d23h4m   10.63    0.14        12      -      -

    20     -         Up           c------       7d23h2m   10.62    0.14        12      -      -

    -------------------------------------------------------------------------------

    Flags:  tunnel: c=compression m=moderate compression a=aggressive compression

                     A=Auto compression f=fastwrite t=Tapepipelining F=FICON

                     T=TPerf i=IPSec l=IPSec Legacy