RE: SMON internals for 10g or 11g?

Peter,

The subject is too big for anyone. But 99% of the time, the real problem comes down to common sense. In your case, it's file access, or corruption:

ORA-00376: file 1 cannot be read at this time ORA-01110: data file 1:
'/u01/app/oracle/oradata/BABYDB/datafile/o1_mf_system_45xrz1n7_.dbf'

If it tries a few times and works, you need to work with the storage or SA team to check the filesystem.

> c. How do u enumerate all the "LIBRARY OBJECT"? And specifically
> for this one:
>
> LIBRARY OBJECT: object=2377171c
> type=TABL flags=EXS/LOC[0005] pflags=[0000] status=VALD load=0

Dump library cache or query x$kglob to enumerate all. I think the number 2377171c uniquely identifies the library cache object. You can find its details in x$kglob matching kglhdobj with '2377171C' (uppercase "C"!). This object is a table in data dictionary and it exists. No load lock. Not sure about the flag.

> d. Decipher the following - which pertained to the above object:
>
> DATA BLOCKS:
> data# heap pointer status pins change whr
> ----- -------- -------- --------- ---- ------ ---
> 0 23778100 237717d8 I/P/A/-/- 0 NONE 00
> 8 2377193c 237711a0 I/P/A/-/- 0 NONE 00
> 9 23771a18 0 I/-/-/-/- 0 NONE 0c

All objects have heap 0. Cursors have 0 and 6. PL/SQL can have 2 and 4. I guess tables and views can have 8 and 9. I/P/A may be initialized, pinned, allocated.

> e. What lwt, pwt, ltm, ptm, and ref/ind?

I don't know.

> f. What is FOB?

The abbreviation FOB appears in various places, such as "Synchronizes various File Object(FOB) operations" in v$lock_type for FP lock. What follows it is detailed info about the file as appeared in the library cache.

This kind of analysis is unlikely needed for solving a problem. But for curiosity, you can take a look at my notes I keep updating over the years:

http://yong321.freeshell.org/oranotes/processstatedumpnotes.txt

Yong Huang       

--
http://www.freelists.org/webpage/oracle-l