DFM - how to get report/graph for volume space used for last year

[sequeirad]

Using Netapp Operations Manager 4.0.0.6953 on Windows Server 2003

I am trying to get a graph / report of the volume space availability over the entire year of 2011 and am having a hard time trying to do that.

Where do you specify the date range when trying to pull up a custom report?

thanks very much

[adaikkap]

Hi,

     The report are all point in time and show the last sample value. If you like to see the space used for last year use the graph and use the custom date ranges like 1d,1w,1m etc.

If you want to use a specific data range use the dfm graph cli and specify the date. If you need more help pls take a look at this doc.

Storage Capacity Management using OnCommand Operations Manager

If you want to know a list of all doc and TR for OnCommand/DFM pls refer the link below.

OnCommand(DFM) and its related Technical Reports

Regards

adai

[sequeirad]

adaikkappan;

thanks very much for this.

I guess I am having a problem with the CLI syntax.

Can you give me the correct syntax for the following?

Want to find volume-usage-vs-total for VOLUME1 from May 1, 2011 to August 31, 2011

thank you

[adaikkap]

Hi

     Pasting the dfm graph cli help. Hope this  helps you construct the cli.

[root@ ~]# dfm graph help

NAME

    graph -- create graphs of data over time

SYNOPSIS

    dfm graph [ <options> ] <graph-name> <name-or-id-to-graph>

DESCRIPTION

    The graph command generates data over time for a particular

    item in the database.

    Use 'dfm graph' with no arguments to get the list of

    graphs.

    The options are

        -s <start-date>

        -e <end-date>

        -D <date-format>

        -F <output-format>

        -h <height>

        -w <width>

    The <date-format> is a time format string as defined

    by the strftime library routine.  The default date format

    is the one appropriate for your locale.

    The <start-date> is the number of seconds in the past that

    the graph should start; the <end-date> is the number of seconds

    in the future that the graph should end. Use a negative

    value for <end-date> if the graph should stop in the past.

    The <height> and <width> are the height and width of the graph

    image in pixels. These options are applicable to image formats

    only i.e. when <output-format> is specified as png or gif.

    Use a suffix afer the <graph-name> to select a different

    range of valid dates; use suffix

        -1d (the default) for data going back 24-48 hours

        -1w for data going back 1-2 weeks

        -1m for data going back 1-2 months

        -3m for data going back 3-6 months

        -1y for data going back indefinitely

    The values above are ranges because the amount of data

    still available depends on whether the current time is near

    the beginning or the end of a day, week, month, or quarter.

    For example, to see the last 60 days of CPU usage for a

    particular storage system, use the "-3m" suffix, because "-1m"

    may not have all the data you request.  The command is

        $ dfm graph -s 5184000 -e 0 cpu-3m system1

    The <output-format> is the type of format in which the output will

    be generated. Supported output formats are text, html, csv, xls,

    png and gif. The output for html, png and gif is in binary format

    and needs to be redirected to a file with proper extension to open it.

    For example, to see the output of the graph 'volume-usage'

    in gif format, issue the command like below

        $ dfm graph -F gif volume-usage > graph.gif

[root@ ~]#

Pls let me know if you still have difficult in getting the data you want.

Regards

adai

[agireesh]

Hi,

Please use below command to list the volume-usage-vs-total graph data.

" dfm graph -s <start-date> -e <end-date> -D '%a, %d %b %Y %H:%M:%S' volume-usage-vs-total"

Where:

<start-date> is the number of seconds in the past that  the graph should start;

<end-date>: is the number of seconds in the future that the graph should end. Use a negative value for <end-date> if the graph should stop in the past.

In your case you want to list graph data from May 1, 2011 to August 31, 2011, So

start-time will be: 8 (months)*30(days)*24(hours)*3600(secs) + 12(days)*24(hours)*3600(secs)=  21772800 secs, approximate   (if today date is 12 Jan 2012)

End date =  4 (months)*30(days)*24(hours)*3600(secs) + 12(days)*24(hours)*3600(secs)= 11404800 secs (this should be in negative )

so, above command should be

"dfm graph -s 21772800 -e -11404800 -D '%a, %d %b %Y %H:%M:%S' volume-usage-vs-total"

If you want to list the data for VOLUME1 then you need to create a group G1, add the volume under group G1 and then use the below command to view the graph data.

"dfm graph -s 21772800 -e -11404800 -D '%a, %d %b %Y %H:%M:%S' volume-usage-vs-total G1"

Hope this help you.

Regards,

Gireesh

[adaikkap]

HI Gireesh,

     Are you sure that the below cli will work ?

"dfm graph -s 21772800 -e -11404800 -D '%a, %d %b %Y %H:%M:%S' volume-usage-vs-total G1"

Since the data you are trying to access is in the past you will have to use the appropriate suffix in the graph name. Since this data is more than 3m old, one can get it only from year history graph.

The suffix -1y needs to be added to the graph name as follows volume-usage-vs-total-1y in the cli to really work.

The help clearly says what suffix to use depending upon the date range.

Use a suffix afer the <graph-name> to select a different

    range of valid dates; use suffix

        -1d (the default) for data going back 24-48 hours

        -1w for data going back 1-2 weeks

        -1m for data going back 1-2 months

        -3m for data going back 3-6 months

        -1y for data going back indefinitely

Regards

adai

[agireesh]

Hi Adai,

Suffix(-1y) need to add here. Because he want to list the 8 months old data so, -1y need to add to the graph name as follows volume-usage-vs-total-1y in the cli,

So Command should be

"dfm graph -s 21772800 -e -11404800 -D '%a, %d %b %Y %H:%M:%S' volume-usage-vs-total-1y G1"

I missed this point, Thanks Adai for this.

Regards,

Gireesh

[sequeirad]

I tried to run the command but am getting this error...

Error: There is no object named '%b'.

[agireesh]

Can you share the command that you are executing. I think, you missed to add single quotes (' ') for time format.

Time format should be in single quotes  ("").

-D '%a, %d %b %Y %H:%M:%S'

Regards,

Gireesh

[sequeirad]

dfm graph -s 21772800 -e -11404800 -D '%a, %d %b %Y %H:%M:%S' volume-usage-vs-total-1y file_types/space_usage

Yes. I am including the single quotes

[rojas]

Hi,

Sorry, But I just want to know how can I manage the amount of time between samples.

When I got cpu-1w output is showing samples every 60 minutes.

If I did that for cpu-1d it is every 15 minutes.

Do I have any way to control time between samples?

Best and Thanks,

Mauricio

[mbeattie]

Hi There,

for anyone reading along I hope the following example helps you figure out how to work with historical data using the "dfm graph" command (for OnCommand developers, it would be great if you could update the documentation and provide additional examples).

The -s <start-date> and -e <end-date> parameters in the help example are based on the number of seconds (because providing a real date like "2013-07-02" would be way too logical right!)

For example, in the following command:

the -D "%Y-%m-%d %H:%M:%S" formats the results (which should work nicely with SQL server %yyyy%-%mm%-%dd% %hh%-%mm%-%ss%")

The -s 86400 is the number of seconds in the past you want to query the data from (86400 = 1 Day (60 seconds * 60 minutes * 24 hours))

The -e 0 means the end time in seconds you want to query

The "cpu" is the name of the graph (to view available graphs run "dfm graph" which will list the defaults)

The -1d is for one day

The "32563" is the DFM object ID of the storage controller you want to run query the data for (To find the dfm object ID run "dfm report view -F csv controllers" and search for the hostname)

Example:

C:\>dfm graph -F csv -D "%Y-%m-%d %H:%M:%S" -s 86400 -e 0 cpu-1d 32563

"Timestamp","CPU Used"

"2013-07-02 12:00:00","17.6"

"2013-07-02 12:15:00","20.4"

"2013-07-02 12:45:00","16.7"

"2013-07-02 13:00:00","9.6"

"2013-07-02 14:00:00","15.9"

"2013-07-02 15:00:00","17.2"

"2013-07-02 16:00:00","14.4"

"2013-07-02 17:00:00","17.0"

"2013-07-02 18:00:00","20.7"

"2013-07-02 19:00:00","16.9"

"2013-07-02 20:00:00","12.4"

"2013-07-02 21:00:00","8.2"

"2013-07-02 22:00:00","10.1"

"2013-07-02 23:00:00","8.4"

"2013-07-03 00:00:00","8.7"

"2013-07-03 01:00:00","31.8"

"2013-07-03 02:00:00","10.0"

"2013-07-03 03:00:00","8.1"

"2013-07-03 04:00:00","11.2"

"2013-07-03 05:00:00","2.7"

"2013-07-03 06:00:00","5.9"

"2013-07-03 07:00:00","13.0"

"2013-07-03 07:45:00","11.3"

"2013-07-03 08:45:00","14.4"

"2013-07-03 09:45:00","15.0"

"2013-07-03 10:45:00","12.6"

"2013-07-03 11:45:00","15.2"

[PANCHATCHARAMP]

Hi,

Can someone help me to pull the past 6 months all the aggregate usage data for a specific filer using single command?

Thanks,

Prem.