Pattern: READNEXT through VSAM, do some SQL, COMMIT, try to keep browsing and REWRITE the VSAM row.
After COMMIT the next READNEXT blows up. REWRITE also returns 16. Feels like I am fighting syncpoint rules.
You are. DB2 COMMIT issues a CICS syncpoint. Syncpoint ends the browse. You need a fresh STARTBR after COMMIT (usually from the key you saved).
REWRITE needs a prior READ/READNEXT with UPDATE. Plain browse READNEXT does not qualify. READNEXT UPDATE is an RLS thing. Typical pattern: browse, save key, COMMIT, READ UPDATE that key, REWRITE, STARTBR again, skip the row you just did if it comes back first.
Rewrote to that pattern. Painful but stable.
You must be signed in to reply to this thread