Backup in non-enterprise

General
1

Hi team,

I tried to find some docs but couldn’t.
Assuming the backup in non-enterprise works, like, sequentially, and not in parallel, so it’s slower and may possibly break on big data.

Is there some docs and comparison or explanation how it works ?

Regards,
ddorian43

2

Hi @ddorian43,

Thanks for asking, haven’t added the docs yet, but started working on it now. For the CE version, you are correct in that it will be sequential (like a DB data dump tool). Have added an example below, will add something very similar into the docs.

Example

Let’s say you have a myapp.stock_market table, along with some sample data inserted into it as shown in this example.

cqlsh> SELECT * FROM myapp.stock_market;

 stock_symbol | ts                  | current_price
--------------+---------------------+---------------
         GOOG | 2017-10-26 09:00:00 |        972.56
         GOOG | 2017-10-26 10:00:00 |     971.90997
         AAPL | 2017-10-26 09:00:00 |        157.41
         AAPL | 2017-10-26 10:00:00 |           157
           FB | 2017-10-26 09:00:00 |        170.63
           FB | 2017-10-26 10:00:00 |     170.10001

(6 rows)

Create Backup

Export the schema for the myapp keyspace by running the following command (in your bash prompt):

$ cqlsh -e "DESC KEYSPACE myapp" > myapp_schema.cql

Export the data using the following command run from inside cqlsh:

cqlsh> COPY myapp.stock_market TO 'myapp_data.csv' WITH HEADER = TRUE ;
...
6 rows exported to 1 files in 0.110 seconds.

We can check the backups using the following:

$ ls -lh myapp_*
     228B May 29 17:36 myapp_data.csv
     327B May 29 17:37 myapp_schema.cql

Restore Backup

To create the keyspace and tables:

cqlsh> source 'myapp_schema.cql'

You should see the following table:

cqlsh> DESCRIBE TABLE myapp.stock_market;

CREATE TABLE myapp.stock_market (
    stock_symbol text,
    ts text,
    current_price float,
    PRIMARY KEY (stock_symbol, ts)
) WITH CLUSTERING ORDER BY (ts ASC)
    AND default_time_to_live = 0;

To restore the data, do the following:

cqlsh> COPY myapp.stock_market FROM 'myapp_data.csv' WITH HEADER = TRUE ;
...
6 rows imported from 1 files in 0.414 seconds (0 skipped).

To verify the data:

cqlsh> SELECT * FROM myapp.stock_market;

 stock_symbol | ts                  | current_price
--------------+---------------------+---------------
         GOOG | 2017-10-26 09:00:00 |        972.56
         GOOG | 2017-10-26 10:00:00 |     971.90997
         AAPL | 2017-10-26 09:00:00 |        157.41
         AAPL | 2017-10-26 10:00:00 |           157
           FB | 2017-10-26 09:00:00 |        170.63
           FB | 2017-10-26 10:00:00 |     170.10001

(6 rows)
1 Like
4

Just an FYI folks - with the recent YugaByte DB change to become 100% open source, all the enterprise including distributed backups referenced here are now available in the open source edition.

Its now just a matter of documenting these - as always, any help in trying out and documenting this feature is most welcome :wink: