Hey Dick, Continuing our discussion I know this is more of a shop specific qns but still do you think that information is stored somewhere. Regards Neelesh. Hello, Database modifications could be tracked not in real-time, though but i've never heard of anything that provides the "last request" info.
One could conceivably write code that intercepted every call to CBLTDLI, recordrd the request and then recorded the response from ims before returning to the actuall caller. Not a trivial task to intercept and inter-leave all of the activity into and back from the database.
How would this be part of converting to db2? Keep in mind that thousands upon thousands of applications have been converted to db2 with no need for something like this. Method Validation - Repeatability. Analytical method validation. Consumer Durables Sector Reports November Related Books Free with a 30 day trial from Scribd. Uncommon Carriers John McPhee. The Art of War Sun Tsu. Related Audiobooks Free with a 30 day trial from Scribd. Elizabeth Howell.
Zaher Nourredine. John Hardy. Views Total views. Actions Shares. No notes for slide. No Changes for several years. I want to update! They were looking at the wholesale conversion of their IMS data to Db2. At one time their IT team had experimented with a manual conversion. It involved only a very small portion of their data, but it took six months to do.
Additional processing is required for transaction threads. Transaction threads are created from a dependent region that was scheduled by the control region. The procedure for transaction thread creation and its use is shown in Figure The commit is done in two phases to synchronize the updates made to IMS databases with those made to DB2 tables. Each participant performs the needed tasks to ensure that a commit is possible for that environment.
DB2 updates its log, retains all locks, and informs the IMS syncpoint coordinator that phase 1 has been completed successfully. If all other participants signal that the commit can proceed, phase 2 is initiated, whereby each participant is responsible for completing the commit. If any participant signals that phase 1 cannot be completed successfully, the entire unit of work is aborted and the updates are backed out.
In phase 2, DB2 logs the commit and releases all locks. In a connection failure of a system crash, however, the commit status of some transactions may be in doubt. They are referred to as in-doubt threads. XRST reads the last checkpoint from the IMS log and passes the data stored in the checkpoint area to the program issuing the command.
The program can use that information to reposition DB2 cursors and reestablish IMS database positioning. It is imperative, though, that each checkpoint call passes all requisite information for repositioning each time it is issued.
For DB2 cursors, this information should include the name of the cursor, the tables being accessed, and the last key or keys retrieved. For IMS databases, this information includes the name of the database, the segment being accessed, and the complete concatenated key. DB2 locks and IMS locks are managed independently. As a result, a deadlock can occur. This deadlock must be resolved before either program can perform subsequent processing. One of the two programs must be targeted as the victim of the deadlock; in other words, it either abends or is timed out.
The deadlock situation is resolved differently depending on the program and the resource. When an MPP is the victim in a deadlock, it abends with a U abend. When a batch-oriented BMP is the victim in a deadlock, the abend received depends on the type of resource that could not be locked:.
When multiple transactions are processed during the same PSB schedule, DB2 can reuse the thread, thereby reducing overhead by avoiding thread creation. The thread is not terminated unless the WFI or Fast Path region is stopped , so these regions tie up a thread indefinitely.
0コメント