Disk Queue Length ?

Our organization is having some slowness problems particularly when most are logging on and off so
mornings and 330 or so I've been through everything bandwidth etc we have 10G switches but I've come across this I believe is the problem on our server that we redirect everyones desktop and profile etc. On that drive in he resource monitor there is a section
for Disk Queue Length that I've read should be 0-2. Ours averages 5-10 and spikes to 50 during these slowness times. All our servers are VMware, its on a SAN with SSD drives so what can I do to resolve this. Its just on the drive that that data is on so we've
been considering creating another drive and splitting up the users profile folders or do we need another separate server? How can I fix this problem? Is there a limit to the amount of users that can be setup to access one server? Do I need to break that up
to several servers?
Jason

Hi Jason0923,
The Disk Queue has Length may caused may reasons, such as high workload with SAN IO bottleneck, generally we can first confirm whether your SAN write disk cache has enabled,
others clue is you can refer the following article to determine whether there have IO bottle neck with your SAN.
Monitoring Queue Length
https://technet.microsoft.com/en-us/library/cc938625.aspx?f=255&MSPPError=-2147217396
Windows Performance Monitor Disk Counters Explained
http://blogs.technet.com/b/askcore/archive/2012/03/16/windows-performance-monitor-disk-counters-explained.aspx
I’m glad to be of help to you!
Please remember to mark the replies as answers if they help and unmark them if they provide no help. If you have feedback for TechNet Support, contact [email protected]

Similar Messages

  • Need help on explanation of Avg. Disk Queue Length

    Based on perfmon, my Avg. Disk Queue Length on physical dick hit 100%.
    What's that mean? Really need explanation

    I’m a bit confused by your statement.   I'm not sure where the 100% is coming from. 
    Avg. Disk Queue Length is the average number of both read and write requests that were queued for the selected disk during the sample interval.
    Current Disk Queue Length is the number of requests outstanding on the disk at the time the performance data is collected. It also includes requests in service at the time of the
    collection. This is a instantaneous snapshot, not an average over the time interval. Multi-spindle disk devices can have multiple requests that are active at one time, but other concurrent requests are awaiting service. This counter might reflect a transitory
    high or low queue length, but if there is a sustained load on the disk drive, it is likely that this will be consistently high. Requests experience delays proportional to the length of this queue minus the number of spindles on the disks. For good performance,
    this difference should average less than two.
    This whole topic can get very confusing.
    Think of Current Disk Queue Length as in flight operations. 
    These are disk read or write that have passes through the Performance Filter Driver and are on their way to the physical disk and back. While in flight a disk activity must pass through (Assuming a SAN) your class drivers, multi path drivers, HBA card
    the network fabric, Switches and into the SAN.  Any of which could introduce a bottleneck. 
    Then the acknowledgment of completion must return.
     Think of Avg. Disk Queue Length as disk activities waiting to jump onto the flight.
    So if you have an Ave. Disk Queue Length happening thinks of this as cars backing up on the on ramp to get on to the highway.
    Typically I start disk analysis by looking at:
    Logical Disk\Ave. Disk sec/Read
    Logical Disk\Ave. Disk sec/Write.
    The Queue Length counters are secondary and only used if the latency counters are out of spec.
    Here are some good Blog and tools to use to follow up.
    Taking Your Server's Pulse
    http://technet.microsoft.com/en-us/magazine/2008.08.pulse.aspx?pr=blog
    Performance Analysis of Logs (PAL) Tool
    http://pal.codeplex.com/
    The Case of the Mysterious Black Box
    http://blogs.technet.com/b/clinth/archive/2009/11/18/the-case-of-the-mysterious-black-box-san-analysis-for-beginners.aspx
    Bruce Adamczak
    Bruce Adamczak

  • VM exhibiting 100% disk busy time, large disk queue lengths

    Hi everyone,
    We have a .VHD workload residing on a logical 2 x 136Gb RAID1 mirror pair of disks.
    The .VHD file is 130Gb (with 70Gb of free space)
    The Virtual Machine is running Windows 2008 R2 SP1, 4 cores and 8Gb of RAM and is exhibiting 100% disk busy time and disk queue lengths of anywhere between 14 and 44
    I'm assuming this is because there is virtually no disk space on the logical drive. Ops Mgr 2012 R2 reports high memory pages/sec
    So we backed up the .VHD workload, broke the RAID1 Mirror and inserted 2 x 300Gb  as a RAID1 mirror and restore the .VHD / VM
    The Logical disk has 50% free disk space, however the VM is still exhibiting 100% disk busy time and the above disk queue lengths.
    It is running on a Windows Server 2008 R2 SP1 HP Proliant Server running the Hyper-V role under Server Core
    Any ideas most appreciated.

    Hi,
    The mirror array doesn’t improve the disk performance but only for the disk redundancy, base on my experience some application frequent operate the large small files often
    can use the large disk resource, if you can’t sure the high disk IO cause by the guest vm or host computer, you can use the Resource Monitor first to identify which process handled the high disk resource, then do the further troubleshooting:
    The third party Resource Monitor use example:
    How to use the Resource Monitor in Windows 7 & Windows 8
    http://www.7tutorials.com/how-use-resource-monitor-windows-7
    Hope this helps.
    We
    are trying to better understand customer views on social support experience, so your participation in this
    interview project would be greatly appreciated if you have time.
    Thanks for helping make community forums a great place.

  • DB, Replay Queue length is growing

    Exchange 2013, it's just starting after migration from 2010 to 2013.
    Replay Queue length in specifed passive DBs which is healthy has been growing rapidly in business
    hour, however Copy Queue length is ok.
    And it's not decreasing them at all in business hour, I'm serching for the cause for that,
    MBX server performance, Disk I/O or networok... need help.
    Even if that's night time, it looks that specified DBs on one server it has long replay Queue length logs.

    HI tanale,
    Seems log files are copied to the passive copies of the mailbox databases. But the log files are not replayed to the passive database.
    Please verify "Don't mount this database at startup " check box selected on the database. If yes please
    uncheck it.
    Regards
    Chinthaka Shameera | MCITP: EA | MCSE: M |
    http://howtoexchange.wordpress.com/

  • Monitoring queue lengths

    Does anyone have any advice/scripts for monitoring queue lengths?
    I'd like to be able to monitor the lengths of the queues within my system, ideally
    such that once queueing occurs an alert/message of sorts can be raised.
    So far I have no continously active monitoring of queue lengths, but am relying
    on the average queue length data provided by the pq command, to identify if queuing
    is occuring.
    Relying on the average queue length reported by pq, I don't think is the best
    route to take. Sometimes it provides data that cannot be correct - I get the
    impression that unless it has a reasonably constant flow of requests it isn't
    very accurate.
    I'm assuming what is actually required is some kind of MIB interrogation program,
    is there anyone that uses something like this to monitor queues?
    The average queue length info provided by pq, does need a little data manipulation
    I've discovered to be meaningful, for everyones benefit here's what needs to be
    done:
    The average queue length is the average number of messages in the queue (inclduing
    those being processed) minus one. I don't know the reason for the minus one,
    but it is something to be aware of (particularly for MSSQ sets).
    I subtract the number of servers serving the queue from the average queue length,
    then add the one back on. This gives the average number of requests in the queue
    that are actually waiting to be processed.
    thanks
    Jody

    Just found it. Coherence->Cache->DistributedCacheForMessages->Attributes->Size

  • Capturing/dumping active thread, queue length and throughput to a file

    Hi,
    I would like to capture active thread information, queue length, and throughput
    information to a file for later analysis. This would be similar to setting the
    verbosegc flag at the java startup to dump the heap space memory usage.
    Is there a way of doing this?

    Just like with heap info, you can use weblogic.Admin GET functionality to query
    WebLogic runtime information.
    To see which bean types you need to query and what properties are available you
    can deploy these 2 jsp's on your weblogic:
    http://dima.dhs.org/misc/listMBeans.jsp
    http://dima.dhs.org/misc/showMBean.jsp
    and point your browser to listMBeans.jsp - the rest is self-explanatory.
    Mark Officer <[email protected]> wrote:
    Hi,
    I would like to capture active thread information, queue length, and throughput
    information to a file for later analysis. This would be similar to setting the
    verbosegc flag at the java startup to dump the heap space memory usage.
    Is there a way of doing this?--
    Dimitri

  • Copy Queue Length - All of a sudden one server having communication issues

    We have 4 servers in a DAG (3 at site A and 1 at site B).
    Of the three servers at site A two of them always show 0 copy queue length.  Recently one of the servers started to show a back log and we are seeing the following in the event viewer.  We see this error when this problem server connects to either
    of the other two in the same physical site.
    The log copier was unable to communicate with server 'ABC1'. The copy of database 'DB2\ABC1' is in a disconnected state. The communication error was: An error occurred while communicating with server
    'ABC1'. Error: Unable to read data from the transport connection: An existing connection was forcibly closed by the remote host. The copier will automatically retry after a short delay.
    At night the queue goes back to 0 and we start over again.  Currently the problem server only has passive copies, we moved of active just in case.
    I have tried using the MAPI network to replicate (Different physical NICs and switches), that was just worse.  Also tried deactivating the primary NIC in the team and using the secondary that is connected to a different core switch.
    Any ideas? 

    Hi,
    Basic on your post, I understand that one DAG member always show 0 copy queue length with error “Unable to read data from the transport connection: An existing connection was forcibly closed by the remote host. The copier will automatically retry after a
    short delay”.
    If I misunderstand your concern, please do not hesitate to let me know.
    Please run below command to double check the connectivity between server:
    1. Use netsh int tcp show global.
    2. Use netsh int tcp to set global autotuninglevel=disabled.
    3. Use netsh int tcp to set global chimney=disabled.
    4. Use netsh int tcp to set global rss=disabled.
    Meanwhile, follow below steps:
    1. Please use the Get-DatabaseAvailabilityGroupNetwork cmdlet to check if DAG network is ok.
    2. Run the Update-MailboxDatabaseCopy -Identity xx cmdlet to seed a copy of a database.
    3. Restart the Microsoft Exchange Replication service.
    4. Please ensure that port 64327 is open.
    Thanks
    Please remember to mark the replies as answers if they help, and unmark the answers if they provide no help. If you have feedback for TechNet Support, contact [email protected]
    Allen Wang
    TechNet Community Support

  • HubTransport UnHealty - Total.Shadow.Queue.Length.Above.Threshold.Monitor - What to check?

    Hi To all,
    looking at Exchange 2013 ServerHealth  i have the "HubTransport" in unhealth state related to this
    item:
    Total.Shadow.Queue.Length.Above.Threshold.Monitor
    I cannot find more information about this issue...
    Many thanks for help! :)
    r.

    Hi,
    Found nothing in the public resource, neither.
    Is there any error/warning/information in the event viewer?
    Please also check the detailed error message in the Monitor if it is possible.
    Did this cause other issues, some like mailflow issue etc.?
    If everything going well, I suggest disable the Alert.
    Thanks
    Mavis
    Mavis Huang
    TechNet Community Support

  • Another question about throughtput and Queue Length

    I do not exactly confirm the meanning of throughput and Queue Length in
    the console of performance of weblogic server.
    AnyOne can give me a explanation? Thanks a lot!

    If you look at the ExecuteQueueRuntime MBean
    (you can use these 2 jsp's:
    http://dima.dhs.org/misc/listMBeans.jsp
    http://dima.dhs.org/misc/showMBean.jsp
    or Sun's HTMLAdaptor: http://dima.dhs.org/misc/StartHtmlAdaptor.jsp
    to browse WLS MBeans.
    PendingRequestCurrentCount
         Returns the number of waiting requests in the queue.
    ServicedRequestTotalCount
         Returns the number of requests which have been processed by this queue.
    ExecuteThreadCurrentIdleCount
         Returns the number of idle threads assigned to the queue
    PendingRequestOldestTime
         Returns the time that the longest waiting request was placed in the queue.
    Eric Nie <[email protected]> wrote:
    I do not exactly confirm the meanning of throughput and Queue Length in
    the console of performance of weblogic server.
    AnyOne can give me a explanation? Thanks a lot!--
    Dimitri

  • Exchange Log Shipping Replay queue length monitor

    Hi Guys,
    Can anyone tell me, what king of monitor is Log shipping replay queue length monitor??
    Is it a average threshold monitor or consecutive samples over threshold monitor?
    Thanks

    Hi,
    This monitor is optimized for the CCR scenario and raises an alert if the number of transaction logs waiting to be committed is greater than 15 logs and has been waiting for more than 5 minutes. Therefore, it is a Consecutive Samples over Threshold.
    You can also get the answer from Microsoft Exchange Server 2007 Management Pack Guide document (Page 72)
    http://download.microsoft.com/download/1/E/D/1ED18BCA-B96D-4184-89DB-EDD9A77E5040/OM2007_MP_EX2007_SP1.doc
    Niki Han
    TechNet Community Support

  • Monitor BizTalk Host Queue length and suspended msgs w/SCOM

    First, I hope the BizTalk forum is the right place to ask this. Maybe I should try the SCOM forum as well.
    I'm trying to create two monitors (Not rules, as we want the alert to be automatically healthy when under treshold again and we want to see the status state as well) in SCOM based on performance counters for BizTalk Msgbox Host Queue Length and suspended
    msgs. My question is what I should use as target (class) in SCOM? And can I use "All instances" of the counter or must i create a monitor for each instance (This is a lot of work and not very dynamic)? We want to monitor all the instances/hosts with
    different tresholds, so the first thing I did was to target the "BizTalk Host" class, so I can do overrides to different hosts.
    The problem with this is it will generete a alert for all hosts if one instance is over treshold. I also tried to target the "Run-time role", and this actually works better, but not perfect as i cannot set a treshold for just one instance/host
    then and it will close the alert if any other intance is under treshold.
    Anyone have experiences with SCOM and monitoring Hosts queues and/or suspended msgs as monitors? 
    thank you in advance for all suggestions!

    I would suggest to look into spool table and its size . As per recommendation it  count should not be greater than 3000 per server .
    Its easy way to monitor the performance counter "Message Box:General Counters /Spool size".you can execute one the following SQL in the BizTalk message box database.
    You can have a counter for spool table size and manually you can use below sql query to find out the count.
    SELECT count(*) from SPOOL WITH (NOLOCK)
    SELECT top 1 rows  FROM sys.partitions WHERE object_id = object_id(‘spool’)
    Note :The NOLOCK keyword is important in first query, you don’t want to put any locks in the spool table while measuring the row count.  The second query is the one used by the performance counter “Spool Size” using the stored procedure
    “MsgBoxPerfCounters_GetSpoolSize”
    Reference :http://msdn.microsoft.com/en-us/library/aa561922.aspx
    Thanks
    Abhishek

  • Thread Count Queue Length in Negative and Too many standby Thread

    We are using Weblogic server 9.2.2 with 1 admin server and 4 managed server . Currently in one of the servers I could observe that there are around 77 standby threads.
    Home > Summary of Servers > server1 > Monitoring > Threads > Self-Tuning Thread Pool
    I could see that the "queue length" is negative (-138) and self tuning standby thread count is 77. Large number of threads STANDBY thread persists during the busy time of the business hours where as other servers are fully utilized.
    Is it normal to have negative queue length and so many STANDBY threads? As for JMS queue negative oracle had already acknowledged that it is a bug. Thanks.
    Edited by: 855849 on May 1, 2011 7:19 AM
    Edited by: SilverHawk on May 12, 2011 8:12 AM

    Yesterday an Oracle Consultant acknowlegded that it is a bug. There was a patch issued for Negative count in the JMS queue count and now this. Thanks for the reply by the way.

  • Wls10mp1  thread pool queue length

    This is under admin console -- server -- threads -- monitoring.
    This field : QueueLength has a -ve value some times.
    Doc states :
    QueueLength
    The number of pending requests in the priority queue. This is the total of internal system requests and user requests.
    Console also shows penging user request count as 0.
    What could cause it to be -ve ?
    Thanx,

    Yesterday an Oracle Consultant acknowlegded that it is a bug. There was a patch issued for Negative count in the JMS queue count and now this. Thanks for the reply by the way.

  • Protecting site and reducing default kernel queue length

    Hi!
    This could be more for BEA support, entering this for far too busy client delivery manager.
    In our setup we sometimes 'loose' PROD due to running out of (doubled) iPlanet service threads. After analysis it looked, that by reducing a problematic application (doubled) weblocig server(s) default kernel queue to lowest value of 256 the PROD could be 'protected'.
    The application is controlled by another service and their tech people are worried is it safe to reduce the kernel queue to lower value.
    According to initial testing on lower environments, this leads only that webserver thread is not reserved, due to bounced request. If the queue is full, this results error in logs and immediately returned (failed) request. After requests time out or if requests are served ok, the normal processing seems to take place.
    Is it ok to reduce the kernel queu e for badly behaving applications? Are there alternative solutions recommended?
    The downstream service is reluctant to make the change, because documentation recommends (weblogic 8.1.5) not to mess with the default settings.
    Alternative solution with own webserver(s) for this service has been studied and does not really seem as an option.
    Best regards,
    Pekka

    Hi,
    We need to ensure the data integrity, please consider to design your azure storage queue logic, please have a look at this article:
    http://azure.microsoft.com/en-gb/documentation/articles/storage-dotnet-how-to-use-queues/
    Regards

  • How to measure slow disk i/o impact on video graphics performance???

    The problem is when an application runs in fullscreen with a high FPS and still shows lag/stutter issues in graphics ...
    Windows 7 doesn`t seem to have any tools that report of stuck disk i/o traffic and neither does the graphic card controls (nvidia). The missing information seem to be:
    1) Whether it is because of a slow drive (fx. http://www.youtube.com/watch?v=QF-SBypQBuw); (Solution: buy a faster drive)
    2) Whether it is because of other programs using the same disk - fx. windows pagefile(?); (Solution: move the app to another drive)
    3) Or whether it is because of a layer program - fx like Sandboxie that redirects filesearch etc.; (Solution: run the app outside the layer)
    Is it really necessary to buy and install and configure a new harddrive and then install the app on this drive uncertain of results just to compare to try solving some graphics lag because of disk i/o issues?

    1) the Windows performance index gives a view of the hardware's expected performance the subscores for harddrive tests should allow you to determine if your harddisk is 'slow'
    2) Resource monitor can help to identify processes that have a lot of IO. Most important measurement will be teh disk queue length. To get even more detailed information, you could use perfmon.
    http://blogs.technet.com/b/askcore/archive/2012/02/07/measuring-disk-latency-with-windows-performance-monitor-perfmon.aspx
    3) virtualisation and/or sandboxing always has a performance impact. You should check with the software vendor how to check for/test the performance impact the application has.
    no, it is not neccasary to buy hardware or software to identify a performance issue. note that in some cases is might be easier/cheaper to buy some new hardware of which you are sure it will meet all requirements.
    PS: consider purchasing an SSD if yous suspect IO issues and are willing to spend some: it will be your best hardwareupgarde you did over the last 10 years!!!
    MCP/MCSA/MCTS/MCITP

Maybe you are looking for