-
We are discussing about export/import not copy command.
Tamil
-
So what? Nothing wrong with mentioning other methodologies. If the original question is "how do i move data only as fast as possible", then COPY is a good solution -- particularly if it can be done with multiple sessions simultaneously. I've often found it to be much faster than export/import, and more robust also.
-
Originally posted by tamilselvan
We are discussing about export/import not copy command.
Tamil
I duuno why ppl tend to look for harder methods when we have simple & the best methods
funky...
"I Dont Want To Follow A Path, I would Rather Go Where There Is No Path And Leave A Trail."
"Ego is the worst thing many have, try to overcome it & you will be the best, if not good, person on this earth"
-
Can anybody elaborate how to do export and import simulataneously using unix pipes or whatever ??
Sanjay G.
Oracle Certified Professional 8i, 9i.
"The degree of normality in a database is inversely proportional to that of its DBA"
-
Steps are:
1 On the Source machine (Let me call it SRC) create unix pipe export
$ mknod /dev/p_exp p
mknod is a command in Solaris to create named pipe.
2 On the Target machine create unix pipe for import
$ mknod /dev/p_imp p
3 On the Target run rsh (remote shell) to copy the file from Source in background and send the output to import pipe
$ rsh SRC dd if=/dev/p_exp > /dev/p_imp
4 On the Source, run the export
$ exp / file=/dev/p_exp full=Y
Wait for 30 Seconds.
5 On the Target, run the import
$ imp / file=/dev/p_imp full=Y
Here order of the steps is very important. Also you should have priv to run rsh. You can also use rcp instead of rsh.
Tamil
To Slimdave,
When I saw copy command as suggested by abhask, I did not think Oracle COPY command, I thought unix cp. That is why I said "We are discussing about exp/imp". Sorry to all.
But I still feel COPY command will be slow, because of dblink is involved.
Tamil
-
Originally posted by tamilselvan
[B]Steps are:
1 On the Source machine (Let me call it SRC) create unix pipe export
$ mknod /dev/p_exp p
mknod is a command in Solaris to create named pipe.
2 On the Target machine create unix pipe for import
$ mknod /dev/p_imp p
3 On the Target run rsh (remote shell) to copy the file from Source in background and send the output to import pipe
$ rsh SRC dd if=/dev/p_exp > /dev/p_imp
4 On the Source, run the export
$ exp / file=/dev/p_exp full=Y
Wait for 30 Seconds.
5 On the Target, run the import
$ imp / file=/dev/p_imp full=Y
Here order of the steps is very important. Also you should have priv to run rsh. You can also use rcp instead of rsh.
Tamil
If you have spare cpu cycles, it might be worth compressing the data on the way into p_exp, and uncompressing it on the way out of p_imp
Originally posted by tamilselvan
To Slimdave,
When I saw copy command as suggested by abhask, I did not think Oracle COPY command, I thought unix cp. That is why I said "We are discussing about exp/imp". Sorry to all.
But I still feel COPY command will be slow, because of dblink is involved.
Tamil
Oh OK, but SQL*Plus copy command doesn't use db_links -- the data gets read by sqlplus from the source db and written to the target db through regular SQL*Net connection.
Therefore you really want the sqlplus session to be running on either the source or the target host to avoid doubling-up on your network traffic
-
Originally posted by tamilselvan
Steps are:
1 On the Source machine (Let me call it SRC) create unix pipe export
$ mknod /dev/p_exp p
mknod is a command in Solaris to create named pipe.
2 On the Target machine create unix pipe for import
$ mknod /dev/p_imp p
3 On the Target run rsh (remote shell) to copy the file from Source in background and send the output to import pipe
$ rsh SRC dd if=/dev/p_exp > /dev/p_imp
4 On the Source, run the export
$ exp / file=/dev/p_exp full=Y
Wait for 30 Seconds.
5 On the Target, run the import
$ imp / file=/dev/p_imp full=Y
Here order of the steps is very important. Also you should have priv to run rsh. You can also use rcp instead of rsh.
Tamil
I bow down to the master of the pipes...
Jeff Hunter
-
Originally posted by tamilselvan
But I still feel COPY command will be slow, because of dblink is involved.
Tamil
Tamil :
It will be worth testing of Simultaneous Exp/Imp Vs COPY ( as suggested by Dave, the SQLPLUS session being open on either Source Or Traget Server ) and dont forget keeping Arraysize 5000 and copycommit your choice preferablly < 50..
Abhay.
funky...
"I Dont Want To Follow A Path, I would Rather Go Where There Is No Path And Leave A Trail."
"Ego is the worst thing many have, try to overcome it & you will be the best, if not good, person on this earth"
-
Originally posted by abhaysk
Tamil :
It will be worth testing of Simultaneous Exp/Imp Vs COPY ( as suggested by Dave, the SQLPLUS session being open on either Source Or Traget Server ) and dont forget keeping Arraysize 5000 and copycommit your choice preferablly < 50..
Abhay.
Either way, you're network will be the limiting factor. Unless you are running multiplexed GB adapters or your disks are really slow, I doubt either method has a significant advantage over the other.
Jeff Hunter
-
Originally posted by marist89
Either way, you're network will be the limiting factor. Unless you are running multiplexed GB adapters or your disks are really slow, I doubt either method has a significant advantage over the other.
Well Network will be limiting factor, but does Import do any better Over Insert using COPY will be the question..So isnt it better to perform a test?
PS I have no idea about "multiplexed GB adapters", if you could let me know about it, would be better (ofcourse I can STW, thas different issue)
Abhay.
funky...
"I Dont Want To Follow A Path, I would Rather Go Where There Is No Path And Leave A Trail."
"Ego is the worst thing many have, try to overcome it & you will be the best, if not good, person on this earth"
Posting Permissions
- You may not post new threads
- You may not post replies
- You may not post attachments
- You may not edit your posts
-
Forum Rules
|
Click Here to Expand Forum to Full Width
|