rman allocating channel

I have a RMAN script that I am building to do a logical validation of datafiles
against an oracle 11.1.0.7 DB on Solaris 5.10. I want the script to allocate
the channels. I tried several variations of the format but below but can't appear to get it to work. Can somebody provide me with the correct syntax
or a good example.

Below is my script and the error I am getting.


rman nocatalog target=/ < ALLOCATE CHANNEL t1 DEVICE TYPE disk;
backup blocks all check logical validate datafile $DATAFILE;
release channel t1;
exit;
EOT



RMAN>
RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-00558: error encountered while parsing input commands
RMAN-01009: syntax error: found "identifier": expecting one of: "for"
RMAN-01008: the bad identifier was: t1
RMAN-01007: at line 1 column 18 file: standard input

You should not need to allocate a channel. RMAN should do that for you.

Correct, RMAN does allocate a channel. let me re-phrase.

What I want to do is run the validation in parallel for several datafiles at a time. Therefore I am looking to allocate channels though my script.

In my example below, it can be the case where $DATAFILE=22,23,24,25
therefore I would like to use 4 channels to process this command so things
run in parallel.

Am I missing something or will RMAN automatically take care of this for me?

Thanks

In rman, connect to your target and run this command.

Code:

---

configure device type disk parallelism 4;

---

And it should cause rman to use parallel of four when you backup or validate your database. You are better off not manually allocating channels when you script rman backups. You can use the following command in rman to tell rman where to backup your data.

Code:

---

configure channel device type disk format '${BACKUPDIR}/${BACKUP_FORMAT}';

---

thanks, your suggestion works great and solves my problem. As for the
second suggestion, thats exactly how I am configured.