DataSunrise Security for Amazon Redshift

Database firewall for Amazon Redshift security is a must when you aim to protect databases against outside threat or inside vulnerabilities. Firewall prevents unauthorized access to database and blocks intrusion attempts. It inspects sql queries for improper content and malicious commands and alerts about suspicious database activity in real time.

Tech Info

DataSunrise Database Firewall protects Amazon Redshift databases from both outside attacks and inside threats. It performs deep traffic filtering on application level. Every incoming and outgoing packet is analyzed for compliance with the customized rules set. In case of meeting a forbidden security rule the blocking scenario includes disconnection of a client from DB or closing of the session.

Proxy mode deployment prevents direct communication between client and database. The firewall acts as an intermediary between client and server, it audits and logs queries or blocks execution of the ones violating firewall's security policies.

proxy-01

Smart SQL analysis algorithms provide continuous monitoring of database activity. During initial traffic processing DataSunrise disassembles user queries, database output etc. Then the queries undergo careful analysis and the firewall determines target database objects, schemas, tables names and other important information. Then the firewall logs queries, blocks their execution or obfuscates database output according to predefined security policies.

 

Scanning of traffic along with detection and blocking mechanisms allows to capture SQL injections in real time. The firewall blocks the following SQL Injection techniques:

  • Union Exploitation Technique
  • Boolean Exploitation Technique
  • Out of band Exploitation technique
  • Time delay Exploitation technique
  • Automated Exploitation
  •  

    Rule conditions include the following parameters:

  • type, instance and name of the target database;
  • whether queries came from a certain client application;
  • whether intercepted queries contain certain SQL statements;
  • whether queries contain signs of SQL injection (OR and UNION statements, comments, double queries, constant expression, keyword in comments);
  • whether queries are directed to certain database elements (schemas, tables, columns, stored procedures).
  •  

    Among supported features there are:

  • All user authentication methods: Kerberos, GSSAPI, SSPI, MD5, SHA256, RADIUS, Trust
  • Processing of SSL-encrypted traffic
  • Data analysis and storage of all query types, including specific replication queries and multiple queries
  • Detailed processing of prepared operations, cursors and portals
  • Full analysis and data collection on COPY operation (transferring data from PostgreSQL table to external files and otherwise). Binary and text formats are supported
  • Analysis of asynchronous queries (Pipelining)
  • Converting all PostgreSQL data types, including composite, range and enum types, from binary format to text
  • Processing of DML/DDL operations in compliance with transactions, search paths and administrator queries
  •