After backup can't find directory to .gz on Centos

ASTRAPI · 03-07-2012, 04:32 AM

When i run this:

I login to ssh and from home directory i was run:

Code:

mysqldump -u root -h localhost -pMYPASSHERE database_name > mydatabase-$(date +%d-%m-%Y).sql; gzip mydatabase-$(date +%d-%m-%Y).sql mydatabase-$(date +%d-%m-%Y).gz

And at the end i got this:

Code:

gzip: mydatabase-07-03-2012.gz: No such file or directory

Any ideas?

acid_kewpie · 03-07-2012, 04:35 AM

well it doesn't exist does it? You're trying to create it, right? just remove that bit from the gzip command, it'll automatically give the output file that name.

ASTRAPI · 03-07-2012, 04:51 AM

Can you please adjust it for me?

I didn't understant it

This one is creating:

Code:

mysqldump -u root -h localhost -pMYPASSHERE database_name > mydatabase-$(date +%d-%m-%Y).sql;

the file mydatabase-07-03-2012.sql on home directory and after this this one:

Code:

gzip mydatabase-$(date +%d-%m-%Y).sql mydatabase-$(date +%d-%m-%Y).gz

Is compressing the mydatabase-07-03-2012.sql to mydatabase-07-03-2012.gz

or not?

I just need to compress my database .gz for backup and then when i need it to be able to decompress it on the same folder with extension .sql

acid_kewpie · 03-07-2012, 04:52 AM

IS what I said really that confusing??

gzip mydatabase-$(date +%d-%m-%Y).sql

lithos · 03-07-2012, 05:19 AM

Quote:

Originally Posted by ASTRAPI

When i run this:
...

Code:

mysqldump -u root -h localhost -pMYPASSHERE database_name > mydatabase-$(date +%d-%m-%Y).sql; gzip mydatabase-$(date +%d-%m-%Y).sql mydatabase-$(date +%d-%m-%Y).gz

...

I do it like this:

Code:

mysqldump -u root -h localhost -pMYPASSHERE database_name > mydatabase-$(date +%d-%m-%Y).sql && gzip mydatabase-$(date +%d-%m-%Y).sql

which is different from yours here at
(date +%d-%m-%Y).sql; gzip
I have &&

Quote:

&& lets you do something based on whether the previous command completed successfully

and I don't have extra mydatabase-$(date +%d-%m-%Y).gz

acid_kewpie · 03-07-2012, 05:22 AM

I'd do it by piping directly into gzip, not redirecting to a file. Could possibly fill up the disk with that decompressed dump

lithos · 03-07-2012, 05:29 AM

Quote:

Originally Posted by acid_kewpie

I'd do it by piping directly into gzip, not redirecting to a file. Could possibly fill up the disk with that decompressed dump

Of course, stupid of me,
that would be :

Code:

mysqldump -u root -h localhost -pMYPASSHERE database_name | gzip > mydatabase-$(date +%d-%m-%Y).sql.gzip

ASTRAPI · 03-07-2012, 05:52 AM

Ok it works

But is it ok for big databases to compress on the fly?

And one last question please:

How can i extract it on the same folder as .sql ?

Thank you

acid_kewpie · 03-07-2012, 05:56 AM

you aren't compressing a database, you're compressing wasteful ASCII text. And you're avoiding not only using the disk space in the first place, but the IO costs of writing it all to the disk and then all back off the disk again, which on a big database, can be a big win.

ASTRAPI · 03-07-2012, 06:11 AM

Is it any better way to compress the database for the backup?

acid_kewpie · 03-07-2012, 06:36 AM

what do you want to be better about it?

ASTRAPI · 03-07-2012, 06:55 AM

Maybe smaller size?

acid_kewpie · 03-07-2012, 06:56 AM

gzip -9

ASTRAPI · 03-07-2012, 07:06 AM

Ok thanks