ARC1244:NO BUILD automatically enabled

Backup Archive and Restore (BAR)
All things BAR
Highlighted
Enthusiast

ARC1244:NO BUILD automatically enabled

I am trying to run a copy to refresh QAS data with the production backup. The backup job ran fine. It's at the bottom of this posting. It had 6 streams in the script.

When I run a copy job I get a warning message. I wonder what parameter I am missing in the copy job. TD database version is 16.10 and Spectrum Protect is used.

 

Warning ARC1244:NO BUILD automatically enabled. Run

                     BUILD after all restore jobs.

 

--My copy job

JOB COPY_QAS_REFRESH_2019TEST1

SESSIONS 23

MULTISTREAM 1

STREAM xxxxxxxxx

COMMANDLINE CSNAME=ASCII CHECKSUM=2 DATAENCRYPTION USE4BVLI

TDPID xxxxxxx

USERID arcuser1

SOURCEUID xxxxxxxxxxxxxx

SCRIPT COPY

LOGON $LOGON;

 

COPY DATA TABLES

 

(QASDB.TABLE1)                                           (FROM(PRODDB.TABLE2)),

(QASDB.TABLE2)                                                     (FROM(PRODB.TABLE2         )),

 

                                                                               

               

RELEASE LOCK,

FILE=ARCHIVE;

 

LOGOFF;

ENDSCRIPT

 

02/28/2019 09:51:26 RELEASE LOCK,

02/28/2019 09:51:26 FILE=ARCHIVE;

02/28/2019 09:51:26 UTILITY EVENT NUMBER - 200353

02/28/2019 09:51:27 LOGGED ON   23 SESSIONS

02/28/2019 09:51:27 ARCHIVE MAPPED TO ARCH_FOR_REFRESH02202019_DSTREAM0001_1.

02/28/2019 09:51:36 *** Warning ARC1244:NO BUILD automatically enabled. Run

                     BUILD after all restore jobs.

 

--The backup job

JOB ARCH_FOR_REFRESH02202019
SESSIONS 48
MULTISTREAM 6
STREAM @All
STREAM @All
STREAM @All
STREAM @All
STREAM @All
STREAM @All
COMMANDLINE CSNAME=ASCII  DATAENCRYPTION
CATALOG ON
TDPID xxxxxx
USERID user1
OTB Tivoli
TIVPOLICY teradata
TIVPOLICY teradata
TIVPOLICY teradata
TIVPOLICY teradata
TIVPOLICY teradata
TIVPOLICY teradata
SCRIPT Archive
LOGON $LOGON;

ARCHIVE DATA TABLES

 

.

.

.

.

 


Accepted Solutions
Teradata Employee

Re: ARC1244:NO BUILD automatically enabled

Each "stream" generates a separate backup dataset - so if there are multiple streams, a single logical "backup image" is split across multiple files (possibly on multiple backup devices).

With ARC, this is called a "cluster" archive.

 

To copy the complete image, you will need multiple steps:

COPY DICTIONARY TABLES to create empty tables in "being restored" state.

COPY DATA TABLES (with NO BUILD stated or implied) for each of the streams / each backup dataset. These need to run in sequence when not restoring to the original configuration, since the data potentially must be redistributed across the AMPs so each job will lock the entire table.

BUILD DATA TABLES to finish the process and make the tables accesible.

1 ACCEPTED SOLUTION
7 REPLIES 7
Teradata Employee

Re: ARC1244:NO BUILD automatically enabled

The backup was split across 6 "streams" but you are only restoring 1 of them in the COPY job. ARC automatically sets NO BUILD in this case, anticipating that you will need to run more partial COPY jobs in sequence, then manually append a BUILD command to the last stream or else run BUILD as a stand-alone job afterwards..

Enthusiast

Re: ARC1244:NO BUILD automatically enabled

I added the same number of streams in the copy job and don't see the warning message. When I run the job it halts. I wonder what's wrong with the script.

 

Thank you,

 

JOB COPY_QAS_REFRESH_2019TEST5

SESSIONS 36

MULTISTREAM 6

STREAM @All

STREAM @All

STREAM @All

STREAM @All

STREAM @All

STREAM @All

COMMANDLINE CSNAME=ASCII  CHECKSUM=2 DATAENCRYPTION USE4BVLI

TDPID xxxxxxx

USERID arcuser1

SOURCEUID xxxxxxxxxxxxxx

SCRIPT COPY

LOGON $LOGON;

 

COPY DATA TABLES

.

.

.

 

RELEASE LOCK,

FILE=ARCHIVE;

 

LOGOFF;

ENDSCRIPT

 

 

 

 

02/28/2019 13:20:06  READING DEFAULTS FROM FILE=[xxxxxxxxxxxxxxxx]

02/28/2019 13:20:06 

02/28/2019 13:20:06   Copyright 1989-2017, Teradata Corporation.

02/28/2019 13:20:06   All Rights Reserved.

02/28/2019 13:20:06 

02/28/2019 13:20:06     ***   ****    ****

02/28/2019 13:20:06    *   *  *   *  *        PROGRAM: ARCMAIN

02/28/2019 13:20:06    *****  ****   *        RELEASE: 16.20.00.03

02/28/2019 13:20:06    *   *  *  *   *        BUILD:   162010eW64 (Dec 22 2017)

02/28/2019 13:20:06    *   *  *   *   ****   

02/28/2019 13:20:06 

02/28/2019 13:20:06  RESTARTLOG = xxxxxxxxxxxxxxxxx1.RLG

02/28/2019 13:20:06  

02/28/2019 13:20:10  PARAMETERS IN USE:

02/28/2019 13:20:10  

 

02/28/2019 13:20:14  RELEASE LOCK,

02/28/2019 13:20:14  FILE=ARCHIVE;

02/28/2019 13:20:14  UTILITY EVENT NUMBER  - 200362

02/28/2019 13:20:16  LOGGED ON   36 SESSIONS

02/28/2019 13:20:18  ARCHIVE MAPPED TO ARCH_FOR_REFRESH02202019_DSTREAM0001_1.

 

 

Enthusiast

Re: ARC1244:NO BUILD automatically enabled

My company PROD system has 7 nodes(one of them is a standby) and DEV 1 node.

When I run a backup with one stream in PRD and a copy job with one stream I have no problem.

 

When I run a backup job with 6 streams in PRD and a copy job with only one stream in DEV the job halts or generates a warning that Build should run.

 

In this situation should the copy job in target system run with only one stream?

 

Is it possible to run a copy job in DEV with multiple streams when there is only one node in the target system (DEV)?

 

 

Teradata Employee

Re: ARC1244:NO BUILD automatically enabled

Each "stream" generates a separate backup dataset - so if there are multiple streams, a single logical "backup image" is split across multiple files (possibly on multiple backup devices).

With ARC, this is called a "cluster" archive.

 

To copy the complete image, you will need multiple steps:

COPY DICTIONARY TABLES to create empty tables in "being restored" state.

COPY DATA TABLES (with NO BUILD stated or implied) for each of the streams / each backup dataset. These need to run in sequence when not restoring to the original configuration, since the data potentially must be redistributed across the AMPs so each job will lock the entire table.

BUILD DATA TABLES to finish the process and make the tables accesible.

Enthusiast

Re: ARC1244:NO BUILD automatically enabled

Thank you for your replies.

I created a copy job with 6 streams and it is running now. After the copy is complete I need to run a REBUILD to make the tables accessible.

 

I searched the web to find REBUILD syntax and each one's syntax looks different.

I believe a BUILD can run on database level and table level and database and table levels together in each script. I could run rebuild on a single database or a single table. But when I ran rebuild on multiple databases or tables the second statement failed in each script. Could you take a look at these scripts and see if they are correct?

Thank you,

 

1. BUILD script on database level

SCRIPT BUILD  <----I wonder if the word 'BUILD' is needed here.
LOGON $LOGON;
BUILD DATA TABLES

(DB1) ;
(DB2) ;
(DB3) ;
(DB4) ;
RELEASE LOCK,
FILE=ARCHIVE;    <---I wonder if 'ARCHIVE' is needed here.

 

 

2. Table level with multiple tables

SCRIPT BUILD
LOGON $LOGON;
BUILD DATA TABLES

(DB1.tb1) ;
(DB2.tb2) ;
(DB3.tb3) ;
(DB4.tb4) ;
RELEASE LOCK,
FILE=ARCHIVE;

 

 

3. Database level and table level together in a script

SCRIPT BUILD
LOGON $LOGON;
BUILD DATA TABLES

(DB1) ;
(DB2) ;
(DB3.tb3) ;
(DB4.tb4) ;
RELEASE LOCK,
FILE=ARCHIVE;

 

Enthusiast

Re: ARC1244:NO BUILD automatically enabled

These scripts worked.

 

1. BUILD script on database level

SCRIPT BUILD 
LOGON $LOGON;
BUILD DATA TABLES

(DB1),
(DB2),
(DB3),
(DB4),
RELEASE LOCK;

 

 

2. Table level with multiple tables

SCRIPT BUILD
LOGON $LOGON;
BUILD DATA TABLES

(DB1.tb1),
(DB2.tb2),
(DB3.tb3),
(DB4.tb4),
RELEASE LOCK;

 

 

3. Database level and table level together in a script

SCRIPT BUILD
LOGON $LOGON;
BUILD DATA TABLES

(DB1),
(DB2),
(DB3.tb3),
(DB4.tb4),
RELEASE LOCK;

 

If you use the same number of streams in your copy job I don't think you need to run a REBUILD after the copy job.

 

Teradata Employee

Re: ARC1244:NO BUILD automatically enabled

Unless you are restoring to the same configuration (in particular, same number of AMPs) the streams must actually do copy/restore one at at time followed by a BUILD.

But if you use the same number of "streams" for the COPY as for the ARCHIVE, perhaps TARA can generate the multi-step process under the covers.

 

Note: BUILD (command in ARC) and "rebuild" utility are entirely different things.