Gestion automatica del almacenamiento en disco

[BeaVeritas]

Buenos días,

Tengo definidos dos almacenamientos de 4TB cada uno que  se corresponden con las unidades F y G del windows. 

Se trata de una maquina virtual y sus discos están en un DataStore VNware de 8TB.

He definido los trabajos de copia para almacenarse en "cualquier almacenamiento en disco", con la esperanza de que, cuando el espacio disponible en el primer disco no sea suficiente, continue en el segundo de forma automática.

Pero no hace esto. En cambio el job se queda indefinidamente en alerta "cargando soporte". Lo he dejado hasta 20 horas para asegurar que no fuera una cuestión de tiempo.

El mensaje es que "no hay ningun soporte "agregable" ".

Al final tengo que cancelar el trabajo y gestionar el espacio de forma MANUAL.

Como puedo resolver esto?

Muchas gracias por su ayuda.

Beatriz

[pkh]

BE does not span disk.  You got to make sure that the target disk has sufficient space to accomodate the entire backup set.

You can automate the selection of the disk if you define a disk pool and put all the disks into it.  You can then use BEMCLI to change the disk selection criteria to select the disk with the most free space.

[BeaVeritas]

Hi pkh!

Thanks for you reply.

I would still ask 2 more questions:

1. Is my best option to send the command Set-BEStorageDevicePool -SelectionMethod MostFreeSpaceFirst ? If I understand well, I will NEVER get any job splitted in more than one disk, isn't it?

2. I found that even when in the currently used disk are several expired copies, it doesn't take that space (doesn't delete expired copies). I found a setting to inforce the deletion ON THE EXPIRATION data, which I understand will occur even when the space is not needed. Is there a way to get expired copy-files deleted WHEN the space is NEEDED? What is happening now is:

- backup job freeze "mounting device"

- I manually delete expired copy-files

- backup job continues 

And I would like my intervention was not needed.

Thanks a lot for your help. 

[pkh · Accepted solution]

1) Yes to both.  That is the setting that you should use and BE will not split a backup set over two disk storage.

2) The are a lot of changes to the way DLM operates over the years and I have lost track of when the changes were implemented.  One of the changes is that the grooming cycle is now 1 hour instead of 4 hours in the case of your version which is BE 2014.  This may help in your case.  Also, I vaguely recall that now DLM will start the grooming process when there is insufficient disk space.  I may be wrong here.

BE 2014 is rather old.  You should upgrade to the latest version to take advantages of the improvements made to DLM.