ANBOB™

The ‘L1 validation’ looks like a segment space management issue. Although it’s not documented (yet) v$event_name suggests that the p1 and p2 parameter are the “segheader” and “l1bmb”. L1 BMB – stands for L1 bitmap block – In an ASSM tablespace this block holds the freeness of a range of blocks within an extent. By freeness I mean the percentage of the free space a block has.

the scenario is:
a) Session needs a block with free space and looks at first relevant L1 bitmap block
b) L1 bitmap block points to data block
c) Session acquires data blocks but finds it is full (or can’t be used for other more subtle reasons)
d) Session repeats from (a) a few times.
e) After a few cycles session adds extent to segment and formats blocks from that extent and continues.

BUGs:

V 11204 <– 12.2

Bug 22188219  Excessive “L1 validation” / “latch: cache buffers chains” inserting into very large non-partitioned table

V 11204 <– 19.1

BUG 27986817 – L1 VALIDATION WAITS WITH HIGH CONCURRENT WORKLAOD

Slow extent allocation for partitioned compressed table causing slowdown with enq: tx – contention and l1 validation wait events

打赏

微信扫一扫，打赏作者吧～