Automated Method to Synchronize Pre-existing Virtual Volumes in a PTP Complex
Original Publication Date: 2004-Sep-28
Included in the Prior Art Database: 2004-Sep-28
An IBM Virtual Tape server (subsequently called VTS) is a tape storage subsystem emulating virtual tape drives and virtual volumes. A Peer-to-Peer Complex (subsequently called PTP Complex or PTP) is an IBM storage subsystem which creates two copies of all virtual tapes in two (or more) Virtual Tape Servers, one copy of each volume on each VTS.. When a PTP is established pre-existing data is not automatically copied from one VTS to the other. This invention disclosure describes a method to initiate the copy of any or all volumes such that the VTS subsystems are synchronized automatically.
THIS COPY WAS MADE FROM AN INTERNAL IBM DOCUMENT AND NOT FROM THE PUBLISHED BOOK
TUC820040015 John C Kennel/Tucson/IBM Nils Haustein, Craig Klein, Dan Winarski
Automated Method to Synchronize Pre -existing Virtual Volumes in a PTP Complex
An IBM Virtual Tape server is a tape storage subsystem emulating multiple tape drives via virtual volumes. A Peer-to-Peer Complex is an IBM storage subsystem which creates multiple copies of each virtual volume in two (or more) Virtual Tape Servers. When a PTP is established pre-existing data volumes are not automatically copied from one VTS to the other. Rather than the customer has to initiate and manage this synchronization. We are presenting a method to automatically copy pre-existing volumes from one VTS in a PTP to another - synchronizing the data in a PTP complex. This method can run concurrently with user operations without any system downtime. The method relies on meta data (so called tokens) which is stored for each virtual volume in a VTS.
The method of copying the data automatically is described via the following text and the flowchart in Figure 1.
1. For each VTS which has been attached to a PTP complex, generate a list of all virtual volumes and
associated physical volumes (111) which have not been copied to the other VTS. This is done by
analyzing the mes_flag, which is part of the meta data. The mes_flag is set to a value of "1" if the
according volume only resides on the local VTS and not on the remote VTS; otherwise, it is set to "0".
2. Sort the list by physical volumes such that virtual volumes sharing the same physical volume (112) are grouped together.
3. For a set of virtual volumes that resides on a physical volume, the mes_flag is changed to 0 and the copy_required flag is set to 1 within the meta data to reflect the fact that each volume needs to be copied (113). A value of "1" in the copy_required flag indicates that the associated virtual volume needs to be copied from the local VTS to the remote VTS.
4. The reconciliation process of the VTS will eventually detect the change in the meta data and initiate the copy of the virtual volume(s). After the copy has been done, the channel director will set the copy_required flag accordingly to reflect that the copy was done. Meanwhile the automated program will wait and monitor the copy_required flag to check progress (114). The preferred embodiment is to calculate the wait time as a function of the number of virtual volumes treated in step 3 (113).
5. The automated method will monitor the progression of the copy operation for each virtual volume processed in step 3 by analyzing the meta data information for those virtual volumes. If a majority of the virtual volumes have been copied (115), the next set of virtual volumes (113) will be processed, assuming there are still volumes to be processed (118). If the meta data shows that the majority of virtual volumes have not been copied yet, the method will wait again for a appropriate...