On Thu, 31 Mar 1994 14:41:09 PST Dinese D.C. Christopher said:
>>I know there was talk at SHARE 82 about tape mirroring, is it going
>>to be supported in the near future?
>
>ADSM does not YET support twining of the files as
>they are backed up, (with one file going to one device and
>an automatic copy of the file going to an alternate device)
>this is a future support item.
>
>Dinese Christopher
>ADSM Technical Support
Dinese
Welcome to the world of email support.
After reading the above msg and reply, I thought that I should open
up the discussion a little more.
At the SHARE 82 ADSM discussions on this topic, The ADSM folks talked
about logical twining and not phyiscal twinning. My understanding was
that a site could define more than one storage pool for any level in
the storage hierarchy. For example one could have the following config
|------| |------|
| | | |
| DB1 | | log1 |
| | | |
|------| |------|
|------| |------|
| | | |
| DB2 | | log2 | Mirror copies of DB1 and log1
| | | |
|------| |------|
|------| |------|
| | | |
| stg1 | | stg1b| Level 1 storage group. Files backed
| | | | from a workstation are stored on
| 3380 | | 3390 | stg1 and stg1b. But each has it's own
| | | | size and migration levels.
|------| |------|
| |
| |
| |
\/ \/
|------| |------|
| | | |
| stg2 | | stg2b| Level 2 storage group. Files on
| | | | stg2 are migrated from stg1. Files
| 3480 | | 3480 | on stg2b are migrated from stg1b.
| | | |
|------| |------|
This would be another example:
|------| |------|
| | | |
| DB1 | | log1 |
| | | |
|------| |------|
|------| |------|
| | | |
| DB2 | | log2 | Mirror copies of DB1 and log1
| | | |
|------| |------|
|------|
| |
| stg1 | Level 1 storage group. Files saved
| | from a workstation are stored on
| 3380 | stg1
| |
|------|
| |
| |___________
| |
\/ \/
|------| |------|
| | | |
| stg2 | | stg2b| Level 2 storage group. Files on stg2 &
| | | | stg2b are migrated from stg1. Since
| 3480 | | 3480 | media on stg2 and stg2b can show
| | | | differences in the capacity from volume
|------| |------| to volume, the volumes will not be
physical copies of each other.
--------------------------------------------------------------------
After thinking about this for a little while, I still have some questions
Will one be able to define different media types at one level. That is
could one define 3380 dasd for stg1 and 3480 carts for stg1b.
Will one be able to define more then two storage pools for each level
in the storage hierarchy.
Can one define different collocation settings for storage pools in
the hierarchy. So the storage pool that is going to be used for
offsite storage would be set to no for collocation. That way one
would not need to mount media stored offsite to append the new data.
Of course this would mean that one would have to update the volume
status to READOnly for the last volume to be written in the series.
This would seem to work best, where the hardware used for the stg2b
could be located offsite. Then one could leave collocation set to
yes.
Any other issues with this that I have not expressed?
Thanks Len Boyle (snolen AT vm.sas DOT com)
|
