tag:blogger.com,1999:blog-4516533711330247058.post4092802743075711224..comments2024-03-28T07:32:09.246-07:00Comments on Robert's Db2 blog: DB2 10 for z/OS: What do You Know About High-Performance DBATs?Roberthttp://www.blogger.com/profile/02058625981006623480noreply@blogger.comBlogger6125tag:blogger.com,1999:blog-4516533711330247058.post-16032122069378541272013-12-23T07:44:31.071-08:002013-12-23T07:44:31.071-08:00Thanks for your response. Thanks for your response. GoWrIsHaNkArhttps://www.blogger.com/profile/15196940946066711322noreply@blogger.comtag:blogger.com,1999:blog-4516533711330247058.post-23810155659214586762013-12-20T20:23:52.346-08:002013-12-20T20:23:52.346-08:00Index entries are handled differently from table r...Index entries are handled differently from table rows in this regard. Index entries MUST be located in the correct physical sequence according to key value. If an index defined with the NOT PADDED option has a key that includes a varying-length column, and the varying-length column is updated so that its length increases to the point that the index entry will no longer fit in the page in which it's located, I would think that the index page would be split. Some of the index page's entries would be moved to a free page in the index so that the enlarged key value could be stored as needed in the correct physical sequence.<br /><br />RobertRoberthttps://www.blogger.com/profile/02058625981006623480noreply@blogger.comtag:blogger.com,1999:blog-4516533711330247058.post-77197760158329660022013-12-19T08:46:45.497-08:002013-12-19T08:46:45.497-08:00i Robert,
I have another doubt not with respect ...i Robert, <br /><br />I have another doubt not with respect to this topic. This is regarding the index DEFINED with VARCHAR column(NOT PADDED).<br /><br />In the case of Table Space , NEARINDREF/FARINDREF will be counted when ever we are trying to update the VARCHAR column in the table due to the effect of DB2 imposing the indirect reference to the original DATA.<br /><br />My question, <br />What would happen for the index defined with NOTPADDED which has the same varchar column as in table and when we are trying to update the column? Will it be relocated from the current postion after deleting the existing KEY ID & POINTER? How DB2 process this request on the INDEX?<br /><br /><br />Please advise.<br /><br />Thanks,<br />Gowri ShankarGoWrIsHaNkArhttps://www.blogger.com/profile/15196940946066711322noreply@blogger.comtag:blogger.com,1999:blog-4516533711330247058.post-88459513253855316002013-12-19T08:34:04.532-08:002013-12-19T08:34:04.532-08:00Thanks Robert.Thanks Robert.GoWrIsHaNkArhttps://www.blogger.com/profile/15196940946066711322noreply@blogger.comtag:blogger.com,1999:blog-4516533711330247058.post-59545963611410648252013-12-11T20:24:31.415-08:002013-12-11T20:24:31.415-08:00X-type (i.e., data change-related) locks on rows o...X-type (i.e., data change-related) locks on rows or pages are always released at commit, regardless of the RELEASE option specified when the DB2 package was bound or rebound. If a package is bound with the cursor stability isolation level (as is usually the case), an S-type lock (data read-related) acquired on a page or row will be released when the associated data has been read by the application process (i.e., when the program moves on from the row or page). This is also not affected by the RELEASE option specified when the package was bound or rebound.<br /><br />Robert<br /><br />RobertRoberthttps://www.blogger.com/profile/02058625981006623480noreply@blogger.comtag:blogger.com,1999:blog-4516533711330247058.post-40315714957197535822013-12-11T11:56:14.758-08:002013-12-11T11:56:14.758-08:00Robert,
Thanks For another wonderful post which ...Robert, <br /><br />Thanks For another wonderful post which helps me as usual to understand the concept.<br /><br />I have got a confuson from your statement<br /><br />"(again, we're talking about tablespace locks and EDM pool elements -- retention of page and row locks is not affected by the RELEASE option of BIND PACKAGE). "<br /><br />Will PAGE/ROW level lock retatined even after the command -MODIFY DDF PKGREL(COMMIT)? If so , How could it give us the compatiblity to run DATA MANIPULATION utilities. Please advise.GoWrIsHaNkArhttps://www.blogger.com/profile/15196940946066711322noreply@blogger.com