Thanks Tom for the loop idea.

But the timing 00:08:14.18 means 8 minutes & 14 seconds.And I think sqlplus have this default format of giving hundreth of a second when we do set timing on. So that means I only waited for 8 minutes.

Hi Tom

I seem to be pestering you alot these days.

I have a lock on a table that I am trying to figure out
how to release.(see v$lock output)


Oracle usercode                Oracle SID Serial nbr Object name     Lock type     F_ground  B_groun
------------------------------ ---------- ---------- --------------- ------------- --------- -------
GUI                                     8          3 MAILING         Row-X (SX)              188

SQL> select * from v$lock where sid = 8;

Thu Oct 23                                                                                              
                                                         Exclusive locks

ADDR     KADDR           SID TY        ID1        ID2      LMODE    REQUEST      CTIME      BLOCK
-------- -------- ---------- -- ---------- ---------- ---------- ---------- ---------- ----------
3A330458 3A330524          8 TX     131138     177618          6          0       6346          0
1D8016AC 1D8016C0          8 TM      65920          0          3          0       6361          0
1D0EAA18 1D0EAA28          8 JQ          0      19148          6          0       6524          0


I ran the dbms_lock.sql to create the dbms_lock package.

I then researched the functions used to allocate,request,convert and release locks.

when I run the dbms_lock.allocate_unique function I keep
getting the following errors.

SQL> EXEC SQL EXECUTE
BEGIN SQL EXECUTE; END;

          *
ERROR at line 1:
ORA-06550: line 1, column 11:
PLS-00103: Encountered the symbol "EXECUTE" when expecting one of the following:
%
The symbol "%" was substituted for "EXECUTE" to continue.


SQL> BEGIN
  2     DBMS_LOCK.ALLOCATE_UNIQUE (lockname IN VARCHAR2,
  3                                lockhandle OUT VARCHAR2); 
  4  END; 
  5  /
   DBMS_LOCK.ALLOCATE_UNIQUE (lockname IN VARCHAR2,
                                          *
ERROR at line 2:
ORA-06550: line 2, column 43:
PLS-00103: Encountered the symbol "VARCHAR2" when expecting one of the following:
(
ORA-06550: line 3, column 55:
PLS-00103: Encountered the symbol ";" when expecting one of the following:
return

I have an ora-00054 "Resource Busy with acquire NOWAIT specified" that I am trying to clear(or release) from the 
system)

as always thanks a heap for your time and help.



 

you know Tom...right after I wrote you I ran this Kill Session script I have that I use to monitor sessions.

I saw that 'locked table' session still hanging out there
and thought...heck...I'm gonna type:

alter system kill session '8,3' immediate;

wa-la the session was slated marked for kill...thought
"ok it's doing the rollback thing..I'll wait awhile"...
looked back and the session was gone and the table was
no longer locked.

Like I said before I always seem to make things
harder than they are...but it's how I learn.

I promise not to pester you(at least for awhile) :~)

regards,
your biggest oracle fan

Found this recomendation recently: </code> http://www.chrispoole.co.uk/tips/plsqltip1.htm <code>

Do you have any comments?
Thanks a lot

This is the problem that we are solving by calling SLEEP with much more than 10 minutes:
Audit procedure runs inside of Oracle. Procedure is getting data through webservice calls to some legacy system and then storing data locally in Oracle. Legacy system is available from 6am till 6pm. Audit can run for hours (for really large data sets - even days. Performance is beyond of us - it's all on the legacy system side). We found the easiest way to solve the problem: schedule procedure to run and periodically check current time between webservice calls. When legacy system is unavailable, procedure is just sleeping till next available working window... Please advice if this is a reasonable approach or it can be done smarter.
Thanks a lot!

What if this must be done as one transaction?

To be honest - no. Audit data retrieved from legacy system is completely isolated and is not used till the audit is done. But I see your concern. So - if I understand you right - you'd definitely prefer splitting such a job to multiple transactions.
By the way - it's typical for this webservice call to run for an hour or more...

Hi Tom,

How accurate is dbms_lock.sleep with smaller duration?

Please consider the following test case:

Version: 9.2.0.6

create table tmp_test (log_dt date not null, id number not null);

begin
for i in 1..3*60*60 loop
insert into tmp_test (log_dt,id) values (sysdate,1);
commit;
dbms_lock.sleep(1);
end loop;
end;
/

select *
from (select x.*, (log_dt - prev_log_dt) * 24 * 60 * 60 as seconds
from (select lag(log_dt) over(order by log_dt) as prev_log_dt, t.*
from test.tmp_test t
where t.id = 1) x)
where seconds > 1
order by log_dt desc
/

PREV_LOG_DT LOG_DT ID SECONDS
3/5/2006 9:42:37 PM 3/5/2006 9:42:39 PM 1 2
3/5/2006 9:41:26 PM 3/5/2006 9:41:28 PM 1 2
3/5/2006 9:40:30 PM 3/5/2006 9:40:32 PM 1 2
3/5/2006 9:39:38 PM 3/5/2006 9:39:40 PM 1 2
3/5/2006 9:38:32 PM 3/5/2006 9:38:34 PM 1 2
3/5/2006 9:37:26 PM 3/5/2006 9:37:28 PM 1 2
3/5/2006 9:36:24 PM 3/5/2006 9:36:26 PM 1 2
3/5/2006 9:35:38 PM 3/5/2006 9:35:40 PM 1 2
3/5/2006 9:34:23 PM 3/5/2006 9:34:25 PM 1 2
...

Even I can find,

select *
from (select x.*, (log_dt - prev_log_dt) * 24 * 60 * 60 as seconds
from (select lag(log_dt) over(order by log_dt) as prev_log_dt, t.*
from test.tmp_test t
where t.id = 1) x)
where seconds > 2
order by log_dt desc
/

PREV_LOG_DT LOG_DT ID SECONDS
3/5/2006 8:57:28 PM 3/5/2006 8:57:38 PM 1 10
3/5/2006 8:48:25 PM 3/5/2006 8:48:28 PM 1 3


I had also captured v$session_wait for this session in another table for every 5 seconds. During the 10 seconds gap, I see only the following:

LOG_DT SID SEQ# EVENT P1TEXT P1 P1RAW P2TEXT P2 P2RAW P3TEXT P3 P3RAW WAIT_TIME SECONDS_IN_WAIT STATE
3/5/2006 8:57:40 PM 126 19587 PL/SQL lock timer duration 100 0000000000000064 0 00 0 00 0 2 WAITING
3/5/2006 8:57:35 PM 126 19584 PL/SQL lock timer duration 100 0000000000000064 0 00 0 00 0 26 WAITING
3/5/2006 8:57:30 PM 126 19584 PL/SQL lock timer duration 100 0000000000000064 0 00 0 00 0 21 WAITING
3/5/2006 8:57:25 PM 126 19581 PL/SQL lock timer duration 100 0000000000000064 0 00 0 00 0 16 WAITING

The wait event is always this.

Does this mean there is some issue with the database server?

Thanks always.

Regards,

Albert Nelson A.

Hi Tom,

I agree with the 2 seconds gap.

I was also very much puzzled with seconds in wait. But I am sure that this is the correct session.

Actually the goal of the test is as below:

We have a table that is supposed to be inserted with about 50 records per second by the application. Of late, we see for few seconds (about 30 to 40) there are no records in the table. This gap is seen about 20 times daily. We are trying to figure it out what is the root cause of this. Is the application itself not inserting the data or the application is sending the inserts but the database is not doing it? That's why I created this test case.

Your advice will be very much helpful for us.

Thanks always.

Regards,

Albert Nelson A.

Hi Tom,

There are no error messages in the alert log.

Albert Nelson A.

Here's the testcase (both 9i and 10g):
Login to sqlplus and do:
exec dbms_lock.sleep(100)

Try to cancel it using CtrlC. It's succesful.
Now when the sleep is placed inside the loop like in:
begin
loop
dbms_lock.sleep(100);
end loop;
end;
/

That block cant be interrupted with CtrlC. The problem is that it cant be interrupted using OCI from another thread as well.
Do you know about any workaround?

Hi Tom,

Yes I find this anomaly.

Even some times, the session_wait shows a continuous increase in seconds_in_wait but during that time I can find records inserted. I had also confirmed that the session id is the correct one.

LOG_DT SID SEQ# EVENT P1TEXT P1 P1RAW SECONDS_IN_WAIT STATE
3/5/2006 8:47:38 PM 126 18966 PL/SQL lock timer duration 100 0000000000000064 118 WAITING
3/5/2006 8:47:33 PM 126 18961 PL/SQL lock timer duration 100 0000000000000064 113 WAITING
3/5/2006 8:47:28 PM 126 18956 PL/SQL lock timer duration 100 0000000000000064 108 WAITING
3/5/2006 8:47:23 PM 126 18951 PL/SQL lock timer duration 100 0000000000000064 103 WAITING
3/5/2006 8:47:18 PM 126 18946 PL/SQL lock timer duration 100 0000000000000064 98 WAITING
3/5/2006 8:47:13 PM 126 18942 PL/SQL lock timer duration 100 0000000000000064 93 WAITING
3/5/2006 8:47:08 PM 126 18937 PL/SQL lock timer duration 100 0000000000000064 88 WAITING
3/5/2006 8:47:03 PM 126 18932 PL/SQL lock timer duration 100 0000000000000064 83 WAITING
3/5/2006 8:46:58 PM 126 18927 PL/SQL lock timer duration 100 0000000000000064 78 WAITING
3/5/2006 8:46:53 PM 126 18922 PL/SQL lock timer duration 100 0000000000000064 73 WAITING
3/5/2006 8:46:48 PM 126 18917 PL/SQL lock timer duration 100 0000000000000064 68 WAITING
3/5/2006 8:46:43 PM 126 18912 PL/SQL lock timer duration 100 0000000000000064 63 WAITING
3/5/2006 8:46:38 PM 126 18907 PL/SQL lock timer duration 100 0000000000000064 58 WAITING
3/5/2006 8:46:33 PM 126 18902 PL/SQL lock timer duration 100 0000000000000064 53 WAITING
3/5/2006 8:46:28 PM 126 18897 PL/SQL lock timer duration 100 0000000000000064 48 WAITING
3/5/2006 8:46:23 PM 126 18892 PL/SQL lock timer duration 100 0000000000000064 43 WAITING
3/5/2006 8:46:18 PM 126 18887 PL/SQL lock timer duration 100 0000000000000064 38 WAITING
3/5/2006 8:46:13 PM 126 18882 PL/SQL lock timer duration 100 0000000000000064 33 WAITING
3/5/2006 8:46:08 PM 126 18877 PL/SQL lock timer duration 100 0000000000000064 28 WAITING
3/5/2006 8:46:03 PM 126 18872 PL/SQL lock timer duration 100 0000000000000064 23 WAITING
3/5/2006 8:45:58 PM 126 18867 PL/SQL lock timer duration 100 0000000000000064 18 WAITING
3/5/2006 8:45:53 PM 126 18862 PL/SQL lock timer duration 100 0000000000000064 13 WAITING
3/5/2006 8:45:48 PM 126 18858 PL/SQL lock timer duration 100 0000000000000064 8 WAITING
3/5/2006 8:45:43 PM 126 18853 PL/SQL lock timer duration 100 0000000000000064 3 WAITING

Log_dt id
1 3/5/2006 8:47:38 PM 1
2 3/5/2006 8:47:37 PM 1
3 3/5/2006 8:47:35 PM 1
4 3/5/2006 8:47:34 PM 1
5 3/5/2006 8:47:33 PM 1
6 3/5/2006 8:47:32 PM 1
7 3/5/2006 8:47:31 PM 1
8 3/5/2006 8:47:30 PM 1
9 3/5/2006 8:47:29 PM 1
10 3/5/2006 8:47:28 PM 1
... Deleted for brevity
113 3/5/2006 8:45:44 PM 1
114 3/5/2006 8:45:43 PM 1

Time to call support?

Regards,

Albert Nelson A.

Hi Tom,

Actually I don't care for the present case. But just that may be a possible bug be pointed out.

Any other suggestions for finding out the reason for gap?

Thanks always.

Regards,

Albert Nelson A.

Hi Tom,

Actually I had captured v$session_wait for all the sessions when I ran the test case.

The application is using connection pool and hence I cannot take identify a particular session.

The application table had seven gaps during that time. I had tried the following query for some gap periods:

select event, count(distinct sid), sum(seconds_in_wait)
from test.sw
where log_dt between
to_date('3/5/2006 8:21:46 PM', 'dd/mm/yyyy hh:mi:ss am') and
to_date('3/5/2006 8:22:49 PM', 'dd/mm/yyyy hh:mi:ss am') and
sid > 11
group by event
/
EVENT COUNT(DISTINCTSID) SUM(SECONDS_IN_WAIT)
PL/SQL lock timer 2 192
SQL*Net message from client 187 6572781
SQL*Net message to client 1 0
buffer busy waits 1 0
db file sequential read 3 0
log file sync 116 0

between
to_date('3/5/2006 8:44:09 PM', 'dd/mm/yyyy hh:mi:ss am') and
to_date('3/5/2006 8:44:59 PM', 'dd/mm/yyyy hh:mi:ss am')

EVENT COUNT(DISTINCTSID) SUM(SECONDS_IN_WAIT)
PL/SQL lock timer 2 620
SQL*Net message from client 201 4688077
SQL*Net message to client 1 0
db file sequential read 1 0
log file sync 126 50

between
to_date('3/5/2006 8:55:03 PM', 'dd/mm/yyyy hh:mi:ss am') and
to_date('3/5/2006 8:55:41 PM', 'dd/mm/yyyy hh:mi:ss am')

EVENT COUNT(DISTINCTSID) SUM(SECONDS_IN_WAIT)
PL/SQL lock timer 2 1656
SQL*Net message from client 221 3646645
SQL*Net message to client 1 0


Does this conclusively prove that application did not sent any insert during that time? Should we concentrate on application / network and rule out database?

Thanks again.

Regards,

Albert Nelson A.

Is there any equivalent SQLPLUS (not pl/sql) command for
the functionality dbms_lock.sleep provides?

dbms_sleep.lock doesnt executes when called
when the database is is in mount state.

Any solution to this?

Can I use the undocumented DBMS_BACKUP_RESTORE.SLEEP?
because it even works in mount state. Will there be any harm?

Actually, on my primary production database I have
set log_archive_duplex_dest='/mydr_archives', /mydr_archives
is a mapped (using Network File System(NFS)) to another
remote AIX machine.

After every 30 minutes archived are generated and stored
on this location.

On the remote AIX machine I am using
RECOVER AUTOMATIC DATABASE USING BACKUP CONTROLFILE UNTIL CANCEL;

But when sometimes it completes before the new file is
arrived, then it requires me to type the command again
when the file is arrived.

What I want is a .SQL script in which I will do
RECOVER AUTOMATIC DATABASE USING BACKUP CONTROLFILE UNTIL CANCEL;
execute dbms_lock.sleep(600);
RECOVER AUTOMATIC DATABASE USING BACKUP CONTROLFILE UNTIL CANCEL;
execute dbms_lock.sleep(600);
.
.
.

I would liked a pl/sql loop, but RECOVER is an SQL command so it will not work and even if it worked, then in case
of its failure due to unavailability of the file it will
break the loop. (by the way, is there a way I can run SQLPLUS command from a pl/sql block.)?

I know you may recommend to implement DATA GUARD, but this
is not an option for us for some reason.

Hi Tom,

I am trying to flip the Standby from MOUNT to OPEN READ ONLY and vice versa through an automated job, so that query reporting can talk to stanby and can get the reports.

However, while trying to open the database in READ ONLY mode I want to make sure that the difference between the applied VS archived log is not more than 1 log sequence.

If the difference is more than 1, please let the process sleep for 5 mins before checking again from v$managed_standby.

I checked dbms_lock.sleep(n) doesn't work and it says Invalid identifier. Can you please let me know any other way by which a process can be made to sleep for n secs before retrying again the operation when in MOUNTED state.

Thanks
VK

Hi Tom,

We are on unix.

However with in PL/SQL code how do I give a host call to let the process sleep for n secs.

declare
Arch_seq NUMBER := 2;
Mrp_seq NUMBER := 0;
begin
-- To check whether we have Applied Vs Archived Status current as of Primary, if not loop through till they are equal-1

While (Arch_seq - Mrp_seq) > 1 loop
for Check_Process in (Select Process, Max(Sequence#) Seq from V\$managed_standby Where Process != 'RFS' Group by Process)
loop
If Check_Process.Process = 'ARCH' then
arch_seq := Check_Process.Seq;
Else
mrp_seq := Check_Process.Seq;
End if;
end loop;
If (Arch_seq - Mrp_seq) > 1 then
dbms_output.put_line('Date : ' || to_Char(sysdate,'dd/mm/yyyy hh24:mi:ss') || ' Waiting for the logs to be applied the difference being ' || (Arch_seq - Mrp_seq) || ' logs');
--dbms_lock.sleep(120);
Else
dbms_output.put_line ('Date : ' || to_Char(sysdate,'dd/mm/yyyy hh24:mi:ss') || ' No Waiting for the logs to be applied');
exit;
End if;
End loop;

Thanks

Hi Tom,

Since, I am not too good at shell scripts, can you please write this for me as it will provide me a good learning platform to move forward in that direction.

Looking for a response.

Thanks

How do you expect to get any better at scripting if you get others to do your work for you?

Not trying to be rude, but personally, I'd rather see Tom spending his time helping people with conceptual problems as opposed to doing their homework.

Regards,

Tyler.

Not to be harsh!

I think you are acting as you are the boss of Tom, let Tom decide what he has to answer and what not.

It's an open community and people do learn from this site a lot, As VK is having this issue I am also facing the same one.

If Tom is busy then you could help us, if you have the knowledge or else you also learn what Tom has to say.

Thanks,

Hi Tom,

I have written the script in flipping the standby from mount to open (RO) and back from open to mount state.

However the logic states that with in a time frame, if the user session exists doing something, we need to kill and put the database back to mount state for the logs to start applying,

However after killing sessions and initiating the recover manage standby database I am getting this error

ORA-01093: ALTER DATABASE CLOSE only permitted with no sessions connected

Can you please let me know how best to handle this?

Hi Tom,

I checked and found that there is an entry which exists in V$session for that sid which has been killed using 'Alter system kill session ' , ' immediate;

I have also set my sqlnext.expire_time=1 which means that the dead connections (which have been killed) should go off with in a span of 1 minute.

However the session does exist for duration of more than 5 mins till I wrote some SQL statement to get 'your session has been killed' and then the session from V$session went off.

Should sqlnet.expire_time not detect dead sessions and remove their entry from V$session.

Thanks

Hello Tom,
The concept of Locks seems enchanting :)
Although i have never used it before as a developer neither know who can use it efficiently.
I wrote this piece of code and did see some changes in the dba_ddl_locks view output (using 10g on vista installed 2 days back)which told me atleast the syntax was correct.However i fail to understand how and where it can be used.we neither specify any object or device when requesting a handle? Or is the handle something like a pointer..that is returned from the package and then we use it...dont know myself i am confused.PLease help clarify.

-- Created on 4/3/2008 by BHUSHAN
declare
-- Local variables here
i integer:=180;
test_lock VARCHAR2(100):= 'var1_test_lock';
test_lock_handle VARCHAR2(100);
begin
-- Test statements here
DELETE test WHERE ROWID IN (SELECT ROWID FROM test WHERE rownum < 100);
sys.dbms_lock.allocate_unique(test_lock,test_lock_handle ,i);
dbms_output.put_line(test_lock_handle);

sys.dbms_lock.sleep(60);
end;

Cheers!!!
Bhushan

Could you please give an example?I have your book but did not carry it with me to Singapore.If you can direct me to any link here we can actually see an example that too would be great..i know the syntax and sementics of how we can get the handle..what i want to know is ocne we get the handle how it can be used...

Cheers!!!
Bhushan

Thanks for the tip re dbms_lock.sleep. My very odd situation combines several of the ideas of this thread.

The problem: an inhouse forms application has a bug where when the user hits the wrong combination of keystrokes a query like this is fired:

select ....
from matched_invoice_line_items W
for update of keyrec_date nowait;

The table has 25 million rows. The W alias seems to be a truncated where clause. This query starts at the beginning of history and locks every row it can, often reading and locking more than 20 million rows before it finds a locked row and must roll back. Generates a lot of redo in a hurry, 3-5gb isn't unusual.

Development hasn't been able to reproduce the bug, Oracle says upgrade your software to supported versions, etc., mgmt has other things to worry about.

The grotesque workaround is to lock the oldest row in the table during working hours, which causes the query-from-hell to fail immediately. The row needs to be locked 13 hours a day.

So in this odd case we have a need for a very long transaction as was mentioned earlier in this thread as well as the need for dbms_lock.sleep - sleep until the window closes and then exit rollback.

It's so ugly that I had to tell someone about it.

Dear Sir,
I have a procedure, which I, have to schedule once every day, But i am not suppose to use dbms_job or dbms_scheduler, so I, am not sure how to create a interval(similar to the one present in dbms_scheduler), so I plan to do something similar to
loop
pl/sql
dbms_lock.sleep(86400);
end loop;
Question 1) Will giving such a long sleep period create issue.
Question 2) Is there any other alternative?
thank you.

Tom,
My assumption is that when you call dbms_lock.sleep that you are actually "doing some work" as far as the session is concerned, you are not idle.

Is this true or not or when you call sleep does Oracle interpret this time as idle ?

Thanks for your response.

Hi Tom,

Would it not be better to leave locking to Oracle itself instead of implementing it using DBM_LOCK? Are there any advantages or is it aimed to be used for specific purposes only.

Thanks

Tom:

I am reading about dbms_lock API from Expert one-on-one Oracle Appendix AG: DBMS_LOCK.

tkyte@TKYTE816> create or replace trigger demo_bifer
2 before insert on demo
3 for each row
4 declare
5 l_lock_id number;
6 resource_busy exception;
7 pragma exception_init( resource_busy, -54 );
8 begin
9 l_lock_id :=
10 dbms_utility.get_hash_value( to_char( :new.x ), 0, 1024 );
11
12 if ( dbms_lock.request
13 ( id => l_lock_id,
14 lockmode => dbms_lock.x_mode,
15 timeout => 0,
16 release_on_commit => TRUE ) = 1 )
17 then
18 raise resource_busy;
19 end if;
20 end;
21 /
Trigger created

I tried the same without using dbms_utility.get_hash_value

rajesh@ORCL> create table t(x number primary key);

Table created.

Elapsed: 00:00:00.40
rajesh@ORCL>
rajesh@ORCL> create or replace trigger t_trig
  2  before insert on T
  3  for each row
  4  declare
  5     l_num number;
  6     resource_busy exception;
  7     pragma exception_init(resource_busy,-54);
  8  begin
  9
 10     if ( dbms_lock.request(
 11                     id => :new.x,
 12                     lockmode => dbms_lock.x_mode,
 13                     timeout =>0,
 14                     release_on_commit=>true
 15                     ) = 1
 16             ) then
 17             raise resource_busy;
 18     end if;
 19  end;
 20  /

Trigger created.

Elapsed: 00:00:00.09
rajesh@ORCL>
rajesh@ORCL> insert into t(x) values (1);

1 row created.

Elapsed: 00:00:00.07
rajesh@ORCL>

And, in the Session-2 I got this

rajesh@ORCL> insert into t(x) values (1);
insert into t(x) values (1)
            *
ERROR at line 1:
ORA-00054: resource busy and acquire with NOWAIT specified or timeout expired
ORA-06512: at "RAJESH.T_TRIG", line 14
ORA-04088: error during execution of trigger 'RAJESH.T_TRIG'


Elapsed: 00:00:00.00
rajesh@ORCL>

Question:

1) Is that dbms_utility.get_hash_value is really needed to get this Non-Blocking fashion?

Hi Guys,
I am Ajay here. I have 3 years of exp in Oracle SQL and PL/SQL.
Please let me know solution for my issue.
I am getting error while compiling procedure in which it is used DBMS_LOCK.sleep (600);

Error(102,10): PLS-00201: identifier 'DBMS_LOCK' must be declared.

create procedure my_sleep( SLEEP_TIME IN NUMBER )
as
begin

for i in 1 .. trunc( SLEEP_TIME/600 )
loop
dbms_lock.sleep( 600 );
end loop;
dbms_lock.sleep( SLEEP_TIME-trunc(SLEEP_TIME/600)*600 );

end;
/
------------------------------------
can we use this proc in the answer for even sleep of 6 to 10 hours without any issue?

Hi :)
Just landed there looking for a sleep in mount.

I came up with using my 1980's C64 BASIC skills

for i in 1 to 10 next i

Sleep 5 seconds on your standby

alter session set PLSQL_OPTIMIZE_LEVEL=0;
for i in 1..1e9 loop null; end loop;

Best regards
Laurent

or simply DBMS_BACKUP_RESTORE.SLEEP(10) ;-)

Note 1296382.1