Uploaded image for project: 'Hadoop Common'
  1. Hadoop Common
  2. HADOOP-10768

Optimize Hadoop RPC encryption performance

    XMLWordPrintableJSON

Details

    • Improvement
    • Status: Patch Available
    • Major
    • Resolution: Unresolved
    • 3.0.0-alpha1
    • None
    • performance, security
    • None
    • Hide
      This improvement leverages AES-NI hardware acceleration to optimize Hadoop RPC performance.

      To enable it, adding the following configuration in core-site.xml
      <property>
      <name>hadoop.rpc.security.crypto.cipher.suites</name>
      <value>AES/CTR/NoPadding</value>
      </property>

      Optionally, set key bit length (default 128) via configuration key hadoop.rpc.security.crypto.cipher.key.bitlength in core-site.xml.
      Show
      This improvement leverages AES-NI hardware acceleration to optimize Hadoop RPC performance. To enable it, adding the following configuration in core-site.xml <property> <name>hadoop.rpc.security.crypto.cipher.suites</name> <value>AES/CTR/NoPadding</value> </property> Optionally, set key bit length (default 128) via configuration key hadoop.rpc.security.crypto.cipher.key.bitlength in core-site.xml.

    Description

      Hadoop RPC encryption is enabled by setting hadoop.rpc.protection to "privacy". It utilized SASL GSSAPI and DIGEST-MD5 mechanisms for secure authentication and data protection. Even GSSAPI supports using AES, but without AES-NI support by default, so the encryption is slow and will become bottleneck.

      After discuss with atm, tucu00 and umamaheswararao, we can do the same optimization as in HDFS-6606. Use AES-NI with more than 20x speedup.

      On the other hand, RPC message is small, but RPC is frequent and there may be lots of RPC calls in one connection, we needs to setup benchmark to see real improvement and then make a trade-off.

      Attachments

        1. Optimize Hadoop RPC encryption performance.pdf
          280 kB
          Dian Fu
        2. HADOOP-10768.011.patch
          99 kB
          Wei-Chiu Chuang
        3. HADOOP-10768.010.patch
          95 kB
          Wei-Chiu Chuang
        4. HADOOP-10768.009.patch
          99 kB
          Wei-Chiu Chuang
        5. HADOOP-10768.008.patch
          94 kB
          Dapeng Sun
        6. HADOOP-10768.007.patch
          94 kB
          Dapeng Sun
        7. HADOOP-10768.006.patch
          95 kB
          Dapeng Sun
        8. HADOOP-10768.005.patch
          94 kB
          Dapeng Sun
        9. HADOOP-10768.004.patch
          121 kB
          Dapeng Sun
        10. HADOOP-10768.003.patch
          78 kB
          Aaron Myers
        11. HADOOP-10768.002.patch
          76 kB
          Dian Fu
        12. HADOOP-10768.001.patch
          71 kB
          Dian Fu
        13. cpu_profile_rpc_encryption_optimize_calculateHMAC.png
          394 kB
          Wei-Chiu Chuang
        14. cpu_profile_RPC_encryption_AES.png
          373 kB
          Wei-Chiu Chuang

        Issue Links

          causes

          Bug - A problem which impairs or prevents the functions of the product. HBASE-20572 HBase2 does not compile against Hadoop3 after HADOOP-10768

          • Blocker - Blocks development and/or testing work, production could not run
          • Open
          duplicates

          Bug - A problem which impairs or prevents the functions of the product. HADOOP-14558 RPC requests on a secure cluster are 10x slower due to expensive encryption and decryption

          • Critical - Crashes, loss of data, severe memory leak.
          • Resolved
          is depended upon by

          Bug - A problem which impairs or prevents the functions of the product. HADOOP-15359 IPC client hang in kerberized cluster due to JDK deadlock

          • Major - Major loss of function.
          • Open
          is duplicated by

          Bug - A problem which impairs or prevents the functions of the product. HADOOP-14558 RPC requests on a secure cluster are 10x slower due to expensive encryption and decryption

          • Critical - Crashes, loss of data, severe memory leak.
          • Resolved
          is related to

          Improvement - An improvement or enhancement to an existing feature or task. HADOOP-12725 RPC encryption benchmark and optimization prototypes

          • Major - Major loss of function.
          • Open
          relates to

          Bug - A problem which impairs or prevents the functions of the product. HBASE-16633 Optimize HBase RPC Encryption Performance

          • Major - Major loss of function.
          • Resolved

          Improvement - An improvement or enhancement to an existing feature or task. HBASE-16414 Improve performance for RPC encryption with Apache Common Crypto

          • Major - Major loss of function.
          • Resolved

          Improvement - An improvement or enhancement to an existing feature or task. HDFS-6606 Optimize HDFS Encrypted Transport performance

          • Major - Major loss of function.
          • Closed

          Improvement - An improvement or enhancement to an existing feature or task. HADOOP-15032 Enable Optimize Hadoop RPC encryption performance for branch-2

          • Major - Major loss of function.
          • Patch Available
          links to

          Web Link Reviews

          Activity

            People

              weichiu Wei-Chiu Chuang
              hitliuyi Yi Liu
              Votes:
              0 Vote for this issue
              Watchers:
              56 Start watching this issue

              Dates

                Created:
                Updated: