Voltage Data Protection w/Java (against CSV and Oracle data sources)

This is an example Java/Maven project demonstrating the use of Voltage SimpleAPI to perform fast and efficient data protection with Format Preserving Encryption (NIST FF1).

Additional binaries needed

See your Voltage Professional Services representative to obtain the SimpleAPI binaries. The following items are not published to public repos, but are needed for execution:

• cache/ — directory to hold the policy cache
• libs/ — external libraries
  • libvibesimplejava.dylib — native dynamic library
  • vibesimplejava.jar — Java archive (import com.voltage.securedata.enterprise)
• trustStore/ — certificates for your Voltage appliance

---

Building the Project (Maven)

You can build this project with the following commands (builds a single fat-jar):

# Installs the Voltage Simple API jar (import com.voltage.securedata.enterprise)
mvn install:install-file -Dfile=libs/vibesimplejava.jar -DgroupId=com.voltage -DartifactId=vibesimplejava -Dversion=1.0 -Dpackaging=jar

# Clean build
mvn clean package

The Java is setup so that passwords and other sensitive info are contained in a ".env" file within the project root (create it; excluded with .gitignore):

ora.user="VOLTAGE"
ora.password="password"
ora.host="hostname.oraclecloud.com"
ora.service="oracle19c.adb.oraclecloud.com"
voltage.secret="sharedsecret"
voltage.host="appliance.voltage.com"
voltage.id="identity"

And run with these commands:

# Run against data.csv as a source
java -Djava.library.path=libs -jar target/protection-1.0.jar "data.csv"

# Run against Oracle as a source
java -Djava.library.path=libs -jar target/protection-1.0.jar --oracle

Optionally use these commands to run without a fat-jar:

java -cp "target/protection-1.0.jar:libs/*" -Djava.library.path=libs com.opentext.voltage.Protection "data.csv"

java --enable-native-access=ALL-UNNAMED -cp "target/protection-1.0.jar:libs/*" -Djava.library.path=libs com.opentext.voltage.Protection "../data.csv"

---

Output

The command line will output:

root@localhost protection % java -Djava.library.path=libs -jar target/protection-1.0.jar --oracle

Exports will save to: "/Voltage/Java/protection/"
Jdbc Url: "jdbc:oracle:thin:@(description= (retry_count=20)(retry_delay=3)(address=(protocol=tcps)(port=1521)(host=<host>))(connect_data=(service_name=<service>))(security=(ssl_server_dn_match=yes)))"
Query: "SELECT * FROM data ORDER BY id"
Total rows ingested = 5000

API Version: Simple API Java 6.22.0.5
Identity: "sampleidentity:070101000000Z:dataprotection.voltage.com#1188515279:data:FPE:256::"
FPE test -->
   Cleartext:  "Hello World"
   Ciphertext: "Eujne Ootlc"
   Cleartext:  "123-45-6789"
   Ciphertext: "762-48-6789"
   Cleartext:  "01-01-2026"
   Ciphertext: "03-23-2007"
   Cleartext:  "1111-2222-3333-4444"
   Ciphertext: "1111-22T5-CFBD-4444"
   Cleartext:  "02/2027"
   Ciphertext: "07/2067"

Dummy enumerator block executed in 3.520125 ms
Empty enumerator block executed in 0.373375 ms

Protect block executed in 26.304667 ms
CSV exported to "/Voltage/Java/protection/protected-ORACLE.data.csv"

Unprotect block executed in 25.874417 ms
CSV exported to "/Voltage/Java/protection/unprotected-ORACLE.data.csv"

---

Speed and execution timing

The code uses a Timer class to calculate code execution time to the nano-second precision. As you can see from above, BLAZING FAST speed.

Timer timer = new Timer();      // Stopwatch instance
timer.reset();  // reset the counter
out("Elapsed time = " + timer.elapsed() + " ms");   // print elapsed time

static class Timer {    // Stopwatch-timer class
    private long start; // holds the start time in nanoseconds
    public Timer() { start = System.nanoTime(); }
    public double elapsed() {   // milliseconds w/nano decimal part
        return (System.nanoTime() - start) / 1_000_000.0;
    }
    public void reset() { start = System.nanoTime(); }
}

---

Oracle SQL

The DDL to make an Oracle data source:

CREATE USER VOLTAGE IDENTIFIED BY "password";

GRANT CREATE SESSION TO VOLTAGE;
GRANT CREATE TABLE TO VOLTAGE;
GRANT UNLIMITED TABLESPACE TO VOLTAGE;

CREATE TABLE VOLTAGE.data (
    id     NUMBER PRIMARY KEY,
    name   VARCHAR2(75),
    ssn    VARCHAR2(11),
    dob    VARCHAR2(10),
    card   VARCHAR2(19),
    exp    VARCHAR2(7)
);

GRANT SELECT, INSERT ON VOLTAGE.data TO VOLTAGE;

SELECT * from VOLTAGE.DATA ORDER BY id

---

Example Data

A "data.csv" is provided in this example. It is also the default value should you not provide any argument to the command at runtime.

id	name	ssn	dob	card	exp
1	Amy Nelson	360-26-1232	09-21-1979	1969-8999-6827-5777	04/2026
2	Betty Hernandez	745-46-3763	09-18-2011	4807-9987-5584-9969	09/2027
3	Fiona Cook	762-85-6426	03-21-1977	5957-4081-1071-9407	07/2026
4	Aiden Martin	895-61-3408	11-02-2018	2905-2293-7691-6411	06/2026
5	Arthur Walker	586-89-8812	08-03-1964	7925-4193-6516-6129	06/2026

After protection, the data will appear as:

id	name	ssn	dob	card	exp
1	Nzr Dtcmsq	328-52-1232	12-29-1963	1969-89LE-K89T-5777	08/1976
2	Pdrfn Tubklaipy	463-57-3763	09-27-1984	4807-99AU-EEYF-9969	08/2066
3	Ziltt Lgkg	119-25-6426	08-26-1981	5957-40EE-V5VL-9407	10/1994
4	Tuucj Vubmhq	531-45-3408	08-01-1953	2905-22KJ-Q4VL-6411	10/2130
5	Cppbvb Dstldn	456-76-8812	02-19-2029	7925-41N7-02GF-6129	10/2130

---

Re-create it!

This data can be re-created by running:

# Generates 5,000 rows of sample data
python3 data-generate.py

And then pushed up to Oracle by:

# Pushes to the DSN in your ".env" file
python3 data-oraclepush.py