Channels and parallelism

DB version: 10.2.0.4

All our items of rman backup are saved in an NFS mount. Our 1 TB database takes about 7 hours for a full backup. We want to increase the performance of RMAN.


will show a SHOW ALL
CONFIGURE DEVICE TYPE DISK PARALLELISM 1 BACKUP TYPE TO BACKUPSET;
CONFIGURE CHANNEL 1 DEVICE TYPE DISK connect 'sys/mypass@thnprod'  FORMAT   '/u04/rmanbkp/thnprod_%U.rmbk';
Question1 .
The Oracle Documentation
http://download.Oracle.com/docs/CD/B19306_01/backup.102/b14191/rcmconc1.htm#i1012864

I think that a channel is actually a server on the target or auxiliary database session and a stream of data to the database for the type of backup device.
We plan to enable parallelism.

If I want to increase the degree of parallelism to 5, I need to set up 5 channels as well. Right?


Question2 .

For the prescription above can I set the configuration that
CONFIGURE DEVICE TYPE DISK PARALLELISM 5 BACKUP TYPE TO BACKUPSET;
CONFIGURE CHANNEL 1 DEVICE TYPE DISK connect 'sys/mypass@thnprod'  FORMAT   '/u04/rmanbkp/thnprod_%U.rmbk';
CONFIGURE CHANNEL 2 DEVICE TYPE DISK connect 'sys/mypass@thnprod'  FORMAT   '/u04/rmanbkp/thnprod_%U.rmbk';
CONFIGURE CHANNEL 3 DEVICE TYPE DISK connect 'sys/mypass@thnprod'  FORMAT   '/u04/rmanbkp/thnprod_%U.rmbk';
CONFIGURE CHANNEL 4 DEVICE TYPE DISK connect 'sys/mypass@thnprod'  FORMAT   '/u04/rmanbkp/thnprod_%U.rmbk';
CONFIGURE CHANNEL 5 DEVICE TYPE DISK connect 'sys/mypass@thnprod'  FORMAT   '/u04/rmanbkp/thnprod_%U.rmbk';
?

10g RMAN automatically opens the number of channels required for your specified parallelism.

That's why

CONFIGURE DEVICE TYPE DISK PARALLELISM 4;

should be enough. (I prefer the numbers 2, 4, 8 for parallelism).

Given that your backups are on a NFS drive, you also suffer network transfer fees.
Why not make compressed backups - that will reduce the size transferred over the network (even if it may take considerably longer if the processors are not good enough. Especially if you have several channels, migh reduce you the number of channels when the compression is activated).

The command is

BACKUP AS COMPRESSED BACKUPSET DATABASE ....

Hemant K Collette

Tags: Database

Similar Questions

  • Problem with SBT and PARALLELISM

    Hello

    I 10.2.0.4 SE on Linux x 86. I do only backups of disk and after I run RMAN I got this error

    RMAN retention policy apply to the order
    RMAN retention policy is set to 7 days recovery window
    using channel ORA_DISK_1
    RMAN-06908: WARNING: operation will not work in parallel on assigned channels
    RMAN-06909: WARNING: parallelism require Enterprise Edition
    output channel: ORA_DISK_1
    RMAN-00571: ===========================================================
    RMAN-00569: = ERROR MESSAGE STACK FOLLOWS =.
    RMAN-00571: ===========================================================
    RMAN-03002: failure of the delete command at 17/02/2010 16:26:31
    ORA-19554: error device allocation, type of device: SBT_TAPE device name:
    ORA-27211: unable to load the media management library
    Additional information: 9


    so... backup is finished on the drive, but it seems that RMAN is to try to assign channels in parallel which I did not need and it is only the functionality of Enterprise Edition still (mine is StandardEdition). Second thing, is he trying to return the SBT_TAPE device, but I need more, I just want the disk.

    Here are rman show all

    RMAN > show all.

    RMAN configuration parameters are:
    CONFIGURE REDUNDANCY 1 RETENTION STRATEGY;
    CONFIGURE BACKUP OPTIMIZATION
    SET UP DEFAULT DISK DEVICE TYPE; # by default
    CONFIGURE CONTROLFILE AUTOBACKUP ON;
    CONFIGURE CONTROLFILE AUTOBACKUP FORMAT FOR DEVICE TYPE DISK TO "%F" # by default
    CONFIGURE CONTROLFILE AUTOBACKUP FORMAT FOR DEVICE TYPE SBT_TAPE TO "%F" # by default
    SET UP THE DEVICE TYPE DISK BACKUP FOR PARALLELISM BACKUPSET TABLET 1 TYPE;
    SET UP THE DEVICE TYPE 'SBT_TAPE' TYPE OF BACKUP FOR PARALLELISM BACKUPSET TABLET 1.
    CONFIGURE BACKUP OF DATA TO DISK FILE TYPE DEVICE TO 1; # by default
    CONFIGURE BACKUP COPIES OF DATA FILE FOR DEVICE TYPE SBT_TAPE TO 1; # by default
    CONFIGURE BACKUP ARCHIVELOG FOR DEVICE TYPE DISK TO 1; # by default
    CONFIGURE BACKUP ARCHIVELOG FOR DEVICE TYPE SBT_TAPE TO 1; # by default
    CONFIGURE CHANNEL DEVICE TYPE DISK MAXPIECESIZE 2 G;
    CONFIGURE CHANNEL DEVICE TYPE 'SBT_TAPE' MAXPIECESIZE 2 G;
    CONFIGURE MAXSETSIZE TO UNLIMITED; # by default
    CONFIGURE ENCRYPTION OF DATABASE # by default
    CONFIGURE THE ENCRYPTION ALGORITHM "AES128"; # by default
    CONFIGURE THE NONE ARCHIVELOG DELETION POLICY; # by default
    CONFIGURE SNAPSHOT CONTROLFILE NAME TO ' / u01/oracle/product/10.2.0/db_1/dbs/snapcf_MSF.f'; # by default


    How to get rid of SBT and PARALLELISM?

    Hi mishomor,

    You must clear all your configurations of rman default no SBT_TAPE.
    (CONFIGURE DEVICE TYPE 'SBT_TAPE' CLEAR;)
    CONFIGURE CHANNEL DEVICE TYPE 'SBT_TAPE' CLEAR ;)

    Also then parallel channel is not allowed as a configuration or in an execution block.

    Can you post the output of show all (after set it up... clear) and after (the part of rman of) a backup of scripts as suggested by Hermant?

    Kind regards

    Tycho

  • Looking through channels and create a chain of search values

    Hello

    I am trying to generate a script that allow me to serch throug a channel and compare the values of a number and create another channel with these instances

    Let me expand

    I have the following test data in a PDM file with a name of group 'Monthly data' and a channel name of "month 2"»

    Time value

    1 18000

    2 45100

    3 59800

    4 35500

    What I want to do is to compare these values to a set of values and record when this value occurred in a channel.

    For eample I am looking for all values greater than 36000 and 50000

    So I would end up with a new channel that would containm the following text

    2 45100

    3 59800

    I know I need to create a table of values that would take the data comparison

    CtrlVals = Array ("", 45000-50000)

    and a channel to store the found values and assign a default value

    IF GroupIndexGet ("stored value") > 0 THEN

    groupIndex = GroupIndexGet ("stored value")

    GroupDel (groupIndex)

    END IF

    Call Autonomecreer ("stored value")

    groupIndex = GroupIndexGet ("stored value")

    CallGroupDefaultSet (GroupIndex)

    And I have to create the channel in that I want to store the data

    ChArray = ChnAlloc ("Instances", iMax): ChnLength (ChArray (0)) = iMax: CtrlEdgeChX = ChArray (0)

    I guess I'm not sure how to insert data found in the created channel.

    I use DIAdem 2011

    Hello SmoothDurban,

    The variable you are looking for is (it would be for the channel of 'Time' in Group 1...

    Data.Root.ChannelGroups (1). Channels ("Time"). Properties ("length"). Value

    You can find the name of a property in DIAdem by dragging in the DataPortal script editor.

    In the above example I have simple glide the maximum property from the data portal to the Panel SCRIPT, the end resulty is the full name of the property with channel and group of references...

    Hope that helps,

    Otmar

  • Find channels and cDAQ modules

    I have looked around the previous messages, but can't seem to find what I'm looking for.

    In most of my experience, I've implemented some DAQ and once the modules and channels are all in place, which is where they stay forever.

    Since I know things will not move, I create tasks in DAQmx.

    I'm working currently on a new installation program, but the problem is that in things cDAQ will not be in the same place every time.

    The application must read a number of strain gauges, which will be connected to the NI 9235 modules.

    It will take place anywhere from 1 to 3 modules attached, and they can be in one of the 8 available cDAQ chassis slots 9172.

    In addition, within a given module, anywhere from 1 to 8 channels will be connected.

    So basically, I guess my questions are running:

    (1) how many modules are connected and where?

    (2) in each module, the number of channels is turned on and how much is empty.

    From there on, I use DAQmx palette to take care of things.

    Thank you!

    So with every new start of the program, you will have a known device (your strain gauge device), but unknown channels. Here's how I attacked only a few years ago...

    • DAQmx create a task to create a job.

    • Make a constant table of all known possible channels. Say that it is 8 cells and compare that to an array of channels USED, making false tracks off the chains 'Unused' and the names of real channel out of used ones. Under certain conditions to check which ones which is used to remove the appropriate channel and add the DAQmx virtual channels properly. The VI snapshot below goes into the details. I brought not only the channels, channels, such as min/max properties and bridge information. My VI also made bridge removal, under certain conditions, depending on sensor that the user wanted to null (that part is not shown.)

    This code isn't super sleek, but I hope this gives you the ideas come from.

  • sample per channel and read sample

    Hello everyone

    I'm new in LABVIEW and I have some difficulties with something.

    I put t know exactly what is the difference between the sample by channel and the sample to read. According to me, knowing that the sample by channel is the size of the buffer that is larger than the sampling frequency, but I put t know what is the sample to read.

    I ve tested with different sample per channel and read. Sometimes I get an error and sometimes know and I would like to know why. If you have examples because I understand better, it will be great.

    I really need to understand this part of my project

    Thanks for your help

    Tony GIBERT

    Hi Roro,

    As you mentioned, when the size of the buffer absorbing continuous samples, you can specify the sample by placing a value to the entry "samples per channel" on schedule vi. The entry of 'number of samples per channel' on reading vi which automatically names a control / constant with 'samples to read' specifies the number of samples you want to remove from the buffer in an appointment during playback of several samples (N). This link may provide a little more detail. I am also attaching a good example of the finder OR example that you may find useful to explore. I assume you are using the DAQmx driver put then please let me know if this isn't the case, but the same principles should apply in any event.

    This means accordingly for sampling at a given rate, you must make sure you're pulling data in large enough 'chunks' so that the buffer overflows (which may well be the cause of error that you see). On the other hand if your sampling rate is slow and your reading vi is having to wait the number of samples to read you have specified to be available, he can lift a time-out error. You can avoid this by increase your sampling rate, reduce your samples to play or increase the time-out specified read vi entry (-1 means it will wait indefinitely).

    Let me know if this helps and how you will.

    All the best.

  • remove the ports series and parallel

    Hello

    I wonder what that "sport erial and parallel can then be removed" in all these articles on the new vSphere 6 functions is all about. I couldn't remove these VMs ports before, couldn't I? Nobody seems to get into the details, everyone just copies / repeat the same thing on the other.

    Best regards

    Jan

    Versions previous hardware could always hardware ports series and parallel within the virtual of the motherboard SuperIO chip, even when there are no virtual serial ports or virtual parallel ports configured for the virtual machine - quietly ports would be "present but disconnected", so they would use more space i/o and IRQ, and always appear to be present in the list of devices in your guest OS IRQ and I/O port map.

    With the latest version of the material, if you delete the port virtual series and virtual parallel port, they will be missing... the guest OS won't see all the device, and it consumes no space port of IO and no IRQ.

    I hope this helps!

    --

    Darius

  • Can I make video tutorails on Adobe creative cloud apps for my youtube channel and website and monetize it

    Can I make video tutorails on Adobe creative cloud applications like photoshop, after effects, first for my youtube channel and website and monetize it

    Here is the official information, if you want to read the details, licenses and terms of use | Adobe

  • The selected file cannot be linked because there 6 audio channels, and the clip was created with 2

    I tried to link the MTS files to a project that I have edited with AVI files. I got the message "the selected file cannot be bound because it has 6 audio channels and clip was created with channel 2/channel" audio. Does anyone know what I need to do?

    I had this problem and the cause in my case, it was that I changed all my clips in the project are mono audio channels. When I tried to swap the originals with proxies, I did, I had a similar message for the clips that have been affected by the designation of channels. I opened the project and selected all my media items in the project window and changed all the audio back to 'use the file' and that did the trick for me.

  • The direct readings and parallel reading

    What is the difference between the direct and parallel reading?

    Direct read:
    http://docs.Oracle.com/CD/E11882_01/server.112/e25513/waitevents003.htm#sthref3915
    http://docs.Oracle.com/CD/E11882_01/server.112/e16638/instance_tune.htm#sthref726

    Db file parallel read:
    http://docs.Oracle.com/CD/E11882_01/server.112/e25513/waitevents003.htm#sthref3901

  • Serial port and parallel to ESXi 4

    Hello

    What about series and PARALLEL to VM mapping in esxi 4?

    Nothing. You cannot map serial ports or parallel in ESXi, ESX is required for this type of functionality.

    ---

    VMware vExpert 2009

    http://blog.vadmin.ru

  • Physical series and parallel Ports

    4 ESXi does physical support Serial and parallel Ports?

    If you consider any comments as useful, please give points

    ---

    VMware vExpert 2009

    http://blog.vadmin.ru

  • Need to find all tables in the target using indicators of loading time APPEND and PARALLEL

    Nice day

    Value your machines Way-Back version 10.1.0.4 for a matter of scheme.

    I am trying to identify all the target tables that are currently defined using APPEND it and PARALLEL loading tips. This is the default setting but is not the optimal method in all cases. With several developers who have developed cards for years, most with little background or training, we try to do a bit of cleanup and optimization.

    With about 900 cards it is not possible to do this manually, so I was hoping that someone who has used this old version remember where this information is stored in the old structure of repository.

    Thank you very much!!

    -gary

    Start OMB * more. And then

    OMBCONNECT...
    OMBCC ' /'
    the value of mods [OMBLIST ORACLE_MODULES]
    foreach $mods {} mod
    OMBCC ' /.
    define tabs [OMBLIST TABLES]
    foreach $tabs {} tab
    puts "$tab [OMBRETRIEVE TABLE '$tab' (PARALLEL_ACCESS_MODE) GET PROPERTIES]"
    }
    playing cards [OMBLIST MAPPINGS]
    foreach map $maps {}
    Set mtabs [OMBRETRIEVE MAPPING "$map" GET of OPERATORS of TABLE]
    foreach mtab $mtabs {}
    puts "$map $mtab [PROPERTIES of OMBRETRIEVE MAPPING"$map"OPERATOR"$mtab"GET (LOADING_HINT)]"
    }
    }
    }

    Kind regards
    Jaap.

  • How to acquire data through several channels in parallel using E 6070 PXI, PXI-4071 and LabVIEW?

    Hello

    I use LabVIEW and NI PXI-4071 PXI NOR 6070E to measure the current through a variable resistance. Now, I use a single channel of SCB - 68, but I want to add another channel at the same time so that I can have two resistors instead of one that I cam measure current through them.

    I have attached a Pdf file showing installation of equipment to use and code LabVIEW also.

    Can someone look at these files and give me some guidelines or ideas that can help me solve this problem, please.

    Thanks in advance.

    Best regards

    Shaheen.

    Your 4071 can do a measure at a time. Your data acquisition cannot measure resistance is not she of the analog inputs.

    However, you could use a multiplexer and multiplexer your 4071 DMM. This habit give you simultaneous action, but can acquire data one after the other, the speed depends on the multiplexer, you choose!

    I hope this helps.

  • channel and sampling rate is not updated until the next cycle

    Hi all

    I'm new to LabVIEW and I wrote the code for the measurement of temperature using the cDAQ-9178 or NI 9214. Could someone please look at my code and help me understand why... my names channel to sample and rate update, until the next time I run my program.

    For example: if I enter the name of the channel "ONE" and "10" sampling frequency... and draw my program will be executed using previous information entered by the user. If I press the race a second time, then it will use the '10' sample rate and channel "ONE". Everyone can't see what I did wrong? I know that my code is absent, but she does everything that I need, except for the update.

    I really want to use a structure of the event, but failed miserably in my attempts. Thank you

    Stream.  Updates the values in your Subvi are run in parallel to the Structures of your event.  The simple solution is to simply put your update of the values inside the event.  In this way the controls are not read until you actually press the next button.

  • Convert the database and parallelism

    Im trying to convert from linux to windows database database. Two of them has a little endian.

    I run the following script in rman and run OK. RMAN is configured to use a parallel process 1 (CONFIGURE DEVICE TYPE DISK PARALLELISM 1 BACKUP BACKUPSET TYPE)
    convert the database new database 'dbsfil '.
    carry the script ' / u05/oracle/dbsfil/transport.sql'
    «Microsoft Windows IA (32-bit)» platform
    format ' / u05/oracle/dbsfil/%U';

    But when I change the property of parallelism Rman to N, it will fail! The error is:

    *******************************
    To change the internal database identifier, use the DBNEWID utility
    RMAN-00571: ===========================================================
    RMAN-00569: = ERROR MESSAGE STACK FOLLOWS =.
    RMAN-00571: ===========================================================
    RMAN-03009: failure of the backup command on the channel ORA_DISK_1 at 2009-07-13 10:43:11
    ORA-19927: CONVERSION of DATABASE Operation cannot continue
    *******************************

    The version of linux database is:
    Oracle Database 10g Enterprise Edition Release 10.2.0.1.0 - Prod
    PL/SQL Release 10.2.0.1.0 - Production
    CORE 10.2.0.1.0 Production
    AMT for Linux: Version 10.2.0.1.0 - Production
    NLSRTL Version 10.2.0.1.0 - Production

    Think you hit a bug 5016125.

    Check metalink Note: ORA-19927 reception when tempting RMAN convert database - 424781.1

Maybe you are looking for