Start a Conversation

Unsolved

This post is more than 5 years old

2329

February 4th, 2014 11:00

Command or Report Needed

Is there a SYMCLI command or combo of commands, or a StorageScope report that will tell me how much storage on a VMAX is being used for TimeFinder, and how much is being used for SRDF ops ?

3 Posts

February 4th, 2014 12:00

Thanks Allen, I did as you said and now have two text files with the Source devices and the SRDF pairs...now wht to do with them next ?

21 Posts

February 4th, 2014 12:00

In storagescope there will be several local replica devices and remote replica devices reports (allocated, raw, etc).  You can refer the online help of storagescope, and check if any of the reports suffice the requirement.  You could also create a custom report in STS if the built in does not give you what you need.

- Brijesh

2.1K Posts

February 4th, 2014 12:00

From SymCLI here is what I would do (off the top of my head). I'm sure there might be easier more elegant ways to do it, but this will at least work as a starting point.

For TimeFinder we don't use BCVs much...most of our TF is Clone. With TF/Clone though you can issue a command agains an array: symclone -sid XXXX list

This will dump you back a list of all active Clone relationships in the array. You could port that to a text file and then edit it to go out and check the capacity of each source device and total them up. If you need help with that we can take that step next.

For SRDF you can run the following: symrdf -sid XXXX list -rdfg all

This will again dump a listing of all RDF pairs on the array (including all meta members). You can dump this out to a text file the same way as the clone list. The extra step here would be to weed out any devices that have an invalid track count listed as "-". These are the meta members. Anything with a "0" or higher will be the meta heads that you want.

Not as easy as you might like, but it's a starting point of how to get there, and once it gets scripted once it can easily be reused as desired.

2.1K Posts

February 4th, 2014 13:00

I would import them into Excel and isolate the column with the device numbers in it (for each one). Then I would add some text before and after that column to build a  "symdev show" command that would also do a "find" on each result and pipe the line found (with the device capacity) out to another text file.

     e.g. symdev -sid XXXX show DEV# | find "MegaBytes" >> textfile.txt

This particular one does have the disadvantage of doublecounting since the symdev show output has two sections where the capacity appears in megabytes, but as long as you know that going in you can then import the resulting text file back into excel yet again and total the capacities.

It ain't pretty, but it gets the job done :-)

465 Posts

February 4th, 2014 15:00

Command:

symdev -sid xx list | findstr /V ".-"

Will produce a list of all user devices, excluding meta members and showing capacity of each device (overall meta size for meta devices). This output also shows RDF attributes.

Command symclone -sid xx list shows all devices involved in clone (at that point int time). It is the target device you are interested in.

Put both into excel and when you have a match of symdevs in the both files, you can determine clone capacity.

You can work out the RDF capacity directly from the symdev output by looking as the attributes.

3 Posts

February 4th, 2014 16:00

Thanks to all ...now I have enough ammo to start a comparison and figure out ht the customer really wants ..that's if they even know :-)

2.1K Posts

February 5th, 2014 06:00

I'm going to add that one to my own bag of tricks Jason. I've never used the symdev list before but I can see where it would be useful for some of the things I've been doing with a repeated symdev show and parsing for specific data.

465 Posts

February 5th, 2014 20:00

symdev list -v will produce symdev show for all devices, so under certain circumstances it can be more efficient to get everything you want in one command and parse from there.

Also, if you asre generating many cli commands in a script, it can put load on the FA CPU's. You can run CLI commands in offline mode against the symapi_db.bin file instead and save all the overhead of going to the FA... faster too!

set environment variables :

SYMCLI_OFFLINE=1

SYMCLI_DB_FILE=

2.1K Posts

February 5th, 2014 21:00

That would work in some environments, but my mixed array with close to 9000 mod3 CKD devices on top of all the FBA devices presents some challenges for dumping a verbose symdev list :-)

I'll keep the offline trick in mind as well, but most of our management hosts have SE on them because they are hosting SMAS or UfV. That could wreak havoc if someone is trying to work through the GUI while I'm forcing SE to work offline. Just need to be careful with that one I guess.

No Events found!

Top