From Fedora Project Wiki

< Infrastructure

Revision as of 21:04, 17 June 2010 by Akistler (talk | contribs) (→‎In Progress: Update wiki formatting - Step 3: Complete)


Points of Contact

Project Sponsor

Name: Mike McGrath
Fedora Account Name: mmcgrath
Group: Infrastructure
Infrastructure Sponsor: mmcgrath

Secondary Contact info

Name: Allen Kistler
Fedora Account Name: akistler
Group: Infrastructure

Name: Huzaifa Sidhpurwala
Fedora Account Name: huzaifas
Group: Infrastructure

Project Info

Project Name: Search Engine Enhancement
Target Audience: All users of Fedora web sites
Expiration/Delivery Date (required): F13

Description/Summary

Fedora needs a search engine[1]

Requirements

  • Crawl the web sites (wiki and non-wiki)
  • Search the web sites (wiki and non-wiki)
  • Java, if any, must be the GCJ/OpenJDK versions in RHEL5; Sun/IBM/BEA Java is not acceptable

Preferences

  • Python-based
  • Programmable keywords to have control over what pages get displayed for certain keywords
  • XML or library interface so other applications can use it
  • Support multiple-language text (i.e., Unicode)

Project Plan

  1. Investigate and evaluate existing open source search engines
  2. Select candidate software
  3. Create public test instances of candidate software
  4. Test for functionality, performance, and impact (re-evaluate, if necessary)
  5. Create capacity and deployment plans
  6. Deploy

Resources Needed

  • Public Test for testing candidate software
  • Permanent home(s) for deployment
    • Web server(s)
    • Database server(s) (maybe)

Software Investigation and Evaluation

Comparison by Requirements

Engine Name Source Language Integrated Crawler Integrated Search Tool Programmable Categories Application Interface
DataparkSearch C Yes Yes Yes
(Tags)
Yes
(Native C API)
Egothor Java
Ferret Ruby
Indri C/C++
KinoSearch Perl/C No
(sample file crawler included)
No
(sample included)
Yes Yes
(BDB/JSON)
mnoGoSearch C Yes Yes Yes
(Tags and Hierarchical categories)
Yes
(Native C API)
Nutch Java Yes
(OpenJDK command line)
Yes
(Tomcat servlet)
No Yes
(Java)
Swish-e C/Perl Yes
(Perl)
Sort Of
(sample included, but has problems)
No
(but can search on META tags)
Yes
(Perl and C APIs)
Xapian C++ Sort Of
(combined Omega with custom Perl)
Yes
(rudimentary Omega CGI)
Yes Yes
(C++, Perl, Python, Ruby)
Engine Name Source Language Integrated Crawler Integrated Search Tool Programmable Categories Application Interface

In Progress

  • DataparkSearch [2] - akistler examined
    • Description
    written in C
    forked from mnoGoSearch in 2003 when mnoGoSearch went semi-commercial
    Indices are stored in a database; Supported databases include MySQL and PostgreSQL (SQLite not advertised, but listed in documentation)
    Support for all SBCS and some DBCS
    Search tool supports Booleans (AND, OR, NOT, NEAR)
    • Evaluation
      • Evaluated after mnoGoSearch.
      • Database modes are single and multi, like mnoGoSearch, but include hashed modes, which does not support string and substring searches, and cached mode, which stores only URI indices in the database, but stores word data in disk files through an additional daemon. The absence of blob mode from mnoGoSearch, but the presence of cached mode in DataparkSearch, appears to be a major difference between the two.
      • As written, the setup expects that the application process logs in to the database as the schema owner, however with additional manual steps it can be made to work as not the schema owner. The call handler setup and custom stored procedure language definition present in mnoGoSearch is commented out in the setup of DataparkSearch, so PostgreSQL superuser is not required, as written. (Their presence in mnoGoSearch is questionable, anyway.) In general, DataparkSearch does appear to be a more slowly developed version of mnoGoSearch.
      • Multiple character set support is not the default, but is specified explicitly before compilation.
      • Bugs in create.multi.sql
        • 185: ERROR: relation "cachedchk2" already exists
        • 186: ERROR: column "url_id" does not exist
        Caused by duplicate CREATE statements in the script. Comment out one set to resolve the bug.
      • Bugs in drop.multi.sql
        • 16: ERROR: sequence "url_rec_id_seq" does not exist
        • 17: ERROR: sequence "categories_rec_id_seq" does not exist
        • 18: ERROR: sequence "qtrack_rec_id_seq" does not exist
        • 19: ERROR: sequence "server" does not exist
        Caused by extraneous DROP statements. Comment them out or ignore them. They're harmless.
      • Extended search mode appears broken. Only the first result is returned. All other results are lost. This might be only an error in the search form, in which case it can be easily debugged and fixed.
    • Requirements
    buildrequires
    • gcc make
    • postgresql-devel (for PostgreSQL support)
    • zlib-devel
    requires
    • postgresql-libs (for PostgreSQL support)
    • httpd
    • zlib
    • others as desired to index documents (pdf, etc.)
    • Setup Notes
      • The tarball was compiled into /opt/dpsearch, although it should be possible to create an RPM for more conventional locations (/bin, /etc, /sbin, etc.)
      • The code compiles and runs fine on x86_64 (compare to mnoGoSearch)
      • Create database user (dp/search) different from database owner (dbowner/dbowner) different from superuser (postgres)
      • Run create.multi.sql as dbowner, then grant privileges to dp
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE cachedchk TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE cachedchk2 TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE categories TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE cookies TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE crossdict TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE dict TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE dict10 TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE dict11 TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE dict12 TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE dict16 TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE dict2 TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE dict3 TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE dict32 TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE dict4 TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE dict5 TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE dict6 TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE dict7 TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE dict8 TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE dict9 TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE links TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE qinfo TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE qtrack TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE robots TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE server TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE srvinfo TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE storedchk TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE url TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE urlinfo TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE categories_rec_id_seq TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE qtrack_rec_id_seq TO dp;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE url_rec_id_seq TO dp;
      • cp /opt/dpsearch/etc/indexer.conf-dist /opt/dpsearch/etc/indexer.conf
      chmod and chgrp to protect passwords!
      • vi /opt/dpsearch/etc/indexer.conf
      set DBAddr, including mode=multi
      set LocalCharset to UTF-8 for full Unicode support
      set MaxHops to the maximum depth (default is 256), if desired/required
      add CrawlDelay, if desired/required
      add "Server Disallow http://fedoraproject.org/w/" and "Server Disallow http://fedoraproject.org/wikiold/"
      set Server to http://fedoraproject.org/ (the URI to be crawled and indexed)
      • cp /opt/dpsearch/etc/stopwords.conf-dist /opt/dpsearch/etc/stopwords.conf
      Note that there are no Chinese stopwords, so don't uncomment the line to configure them
      • cp /opt/dpsearch/etc/langmap.conf-dist /opt/dpsearch/etc/langmap.conf
      • vi /opt/dpsearch/etc/langmap.conf
        • Comment out the lines for ca.latin1.lit.lm, ga.latin1.lit.lm, ko.utf8.lit.lm, and pt-pt.latin1.lm, because the files for them don't exist
        • Uncomment the lines for ja.euc-jp.lm, ja.sjis.lm, ta.tscii.lm, zh.big5.lm, and zh.gb2312.lm, because the files for them do exist
      • cp /opt/dpsearch/etc/sections.conf-dist /opt/dpsearch/etc/sections.conf
      • cp /opt/dpsearch/etc/search.htm-dist /opt/dpsearch/etc/search.htm (htm name must not change; compare to mnoGoSearch)
      chmod and chgrp to protect passwords!
      • vi /opt/dpsearch/etc/dpsearch.htm
      set DBAddr and LocalCharset to correspond to /opt/dpsearch/etc/indexer.conf
  • Egothor [3]
    • Description
    written in Java
  • Ferret [4]
    • Description
    Ruby port of Lucene
    KinoSearch and Ferret intend to merge as Lucy [5]
  • Indri [6]
    • Description
    Written in C/C++
    Indexes text and HTML, can be extended with custom parsers
    Provides Java, PHP, and C++ APIs and, if desired, a SOAP interface
    Search tool is a local Java application or a PHP CGI application
  • KinoSearch [7] - akistler examined
    • Description
    Perl/C port of Lucene
    in Fedora already
    maintainer Rectangular Research appears to be just one guy, who considers KinoSearch to be alpha software
    KinoSearch and Ferret intend to merge as Lucy [5]
    • Evaluation
    Search engine library with sample indexer and search page rather than fully-functional application. Stores indices in Berkeley DB files with JSON interfaces. Allows custom-designed indices, including categories (exact match) to fulfill "programmable keywords" requirement. Each document index on each document source is a single write-once file collection (BDB and JSON) in a unique directory. Rerunning the indexer creates a new directory, obsoleting the old directory if all the old documents are included. The old directory then needs to be cleaned up. Postings can, however, be deleted from an index. Additionally, only the new documents can be indexed, but that's not efficient.
    • Requirements
    buildrequires
    • gcc
    • (EPEL) perl-Module-Build
    requires
    • (EPEL) perl-JSON-XS
    Problem: Desires 1.53, but EPEL has 1.43
    Note: http://web.archive.org/web/20071122035408/search.cpan.org/src/MLEHMANN/JSON-XS-1.53/
    Note: works with 1.43, anyway
    • (EPEL) perl-Lingua-Stem-Snowball
    • (EPEL) perl-Lingua-StopWords
    • (EPEL) perl-Parse-RecDescent
    sample indexer reads files from the file system and requires
    • (EPEL) perl-HTML-Tree
    sample cgi search script requires
    • (CPAN) Data::Pageset (which requires Data::Page)
    • (EPEL) perl-Test-Exception
    • (EPEL) perl-Class-Accessor-Chained
  • mnoGoSearch [8] - akistler examined
    • Description
    written in C
    UNIX/Linux source code is GPL; Windows binaries are commercial, likely based on the GPL UNIX/Linux code, and lag a few versions behind
    Indices are stored in a database; Supported databases include MySQL, PostgreSQL, and SQLite (among others)
    HTTP, FTP, and NNTP crawling
    C, PHP, and Perl APIs (advertised, apparently only C API really included)
    SBCS and most MBCS supported, including most eastern Asian languages
    • Evaluation
      • The supplied install.pl script generates a configure command, but does not support SQLite.
      • Adding --with-sqlite3 to the generated command adds SQLite support. An empty database must be created manually. A URI in the indexer.conf file specifies the location of the database. According to the documentation, sqlite:/path/to/db/file should work, but doesn't. According to the message boards on mnoGoSearch.org, sqlite://localhost/path/to/db/file should work, but doesn't.
      • No problems compiling with PostgreSQL support. Configure and Makefile may need significant rewriting to make good RPMs. There is no support for cross-compiling to 32-bit architectures from 64-bit machines.
      • As written, the setup expects that the application process logs in to the database as a PostgreSQL superuser and schema owner, however with additional manual steps it can be made to work as neither the superuser nor the schema owner. Several schemata are available (single, multi, and blob). LavTech recommends blob for sites indexing more than 50k documents. The crawler is very flexible with quite a complex configuration file. The CGI search page also has nice features for "advanced" searching, although it can be customized to suit each site. Tags are labels configured within the crawler, usually by URI server component. Categories are numerical hierarchies, up to 6 levels deep, also specified in the crawler configuration.
      • Bugs in pgsql/drop.blob.sql
        1. drop function clean_srvinfo(); (the () is omitted, but needs to be included)
        2. DROP LANGUAGE plpgsql; (missing)
        3. DROP FUNCTION plpgsql_call_handler(); (missing, has to be run twice, once for postgres, once for dbowner?)
          Does this definition in postgres (dangerously) shortcut anything inherent for other DBs on the same server?
    • Requirements
    buildrequires
    • gcc make
    • sqlite-devel (for SQLite support)
    • postgresql-devel (for PostgreSQL support)
    • zlib-devel
    requires
    • sqlite (for SQLite support)
    • postgresql-libs (for PostgreSQL support)
    • httpd
    • zlib
    • others as desired to index documents (pdf, etc.)
    • Setup Notes
      • The tarball was compiled into /opt/mnoGoSearch, although it should be possible to create an RPM for more conventional locations (/bin, /etc, /sbin, etc.)
      • On x86_64 architecture, the x86_64 binary fails when indexing a crawl with the error "indexer[21272]: PQexecPrepared: ERROR: incorrect binary data format in bind parameter 2." The tarball refuses to cross-compile to 32-bit architecture, despite tweaking the ./configure options. Compiling on a 32-bit machine and moving the binaries to the 64-bit machine works.
      • Create database user (mno/search) different from database owner (dbowner/dbowner) different from superuser (postgres)
      • /opt/mnoGoSearch/sbin/indexer -Ecreate, to create the tables, etc., and /opt/mnoGoSearch/sbin/indexer -Edrop, to drop tables, etc., just runs a script from /opt/mnoGoSearch/share/<db-type>/, but needs to be run as postgres
      • Run create.blob.sql as postgres, change owners to dbowner, and grant privileges to mno
      ALTER TABLE bdict OWNER TO dbowner;
      ALTER TABLE bdicti OWNER TO dbowner;
      ALTER TABLE categories OWNER TO dbowner;
      ALTER TABLE crossdict OWNER TO dbowner;
      ALTER TABLE dict OWNER TO dbowner;
      ALTER TABLE links OWNER TO dbowner;
      ALTER TABLE qcache OWNER TO dbowner;
      ALTER TABLE qinfo OWNER TO dbowner;
      ALTER TABLE qtrack OWNER TO dbowner;
      ALTER TABLE server OWNER TO dbowner;
      ALTER TABLE srvinfo OWNER TO dbowner;
      ALTER TABLE url OWNER TO dbowner;
      ALTER TABLE urlinfo OWNER TO dbowner;
      ALTER TABLE wrdstat OWNER TO dbowner;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE bdict TO mno;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE bdicti TO mno;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE categories TO mno;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE crossdict TO mno;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE dict TO mno;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE links TO mno;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE qcache TO mno;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE qinfo TO mno;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE qtrack TO mno;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE server TO mno;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE srvinfo TO mno;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE url TO mno;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE urlinfo TO mno;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE wrdstat TO mno;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE categories_rec_id_seq TO mno;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE qtrack_rec_id_seq TO mno;
      GRANT SELECT,INSERT,UPDATE,DELETE ON TABLE url_rec_id_seq TO mno;
      • cp /opt/mnoGoSearch/etc/indexer.conf-dist /opt/mnoGoSearch/etc/indexer.conf
      chmod and chgrp to protect passwords!
      • vi /opt/mnoGoSearch/etc/indexer.conf
      set DBAddr, including mode=blob for > 50k documents
      set LocalCharset to UTF-8 for full Unicode support
      set MaxHops to the maximum depth (default is 256), if desired/required
      add CrawlDelay, if desired/required
      add "Server Disallow http://fedoraproject.org/w/" and "Server Disallow http://fedoraproject.org/wikiold/"
      set Server to http://fedoraproject.org/ (the URI to be crawled and indexed)
      • cp /opt/mnoGoSearch/etc/search.htm-dist /opt/mnoGoSearch/etc/mnoGoSearch.htm (htm name must match cgi name below)
      chmod and chgrp to protect passwords!
      • vi /opt/mnoGoSearch/etc/mnoGoSearch.htm
      set DBAddr and LocalCharset to correspond to /opt/mnoGoSearch/etc/indexer.conf
      • cp /opt/mnoGoSearch/bin/search.cgi to /var/www/cgi-bin/mnoGoSearch
      • To crawl the sites configured, /opt/mnoGoSearch/sbin/indexer
      • To index the data collected, /opt/mnoGoSearch/sbin/indexer -Eblob
      • To display database statistics, /opt/mnoGoSearch/sbin/indexer -S
      • To clear the database, /opt/mnoGoSearch/sbin/indexer -C
  • Nutch [9] - akistler examined
    • Description
    written in Java
    based on Lucene
    the crawler/indexer is a Java command line application; the default depth is 5; the default number of threads is 10
    the search tool runs in a Java servelet container, e.g., Tomcat
    • Evaluation
    There's nothing to build. Simply configure the crawler (which is actually the indexer, too) and deploy/configure the searcher. The crawler caches the pages it indexes, making the cache available to the search tool. The search interface is extremely simple and is multi-lingual, but is almost entirely an advertisement for the Nutch project. It doesn't look particularly easy to rebrand. Overall, the polish of the finished product means it's less flexible to custom modifications, like programmable keywords. After creating a new index (i.e., after a new crawl), the search application must be reloaded in Tomcat manager. The crawler is more flexible than a brief investigation could reveal. The official documentation leaves a lot to be desired. Searches are for single terms only, no multiple terms or +/- Booleans.
    • Requirements
    java-1.6.0-openjdk, java-1.6.0-openjdk-devel (explicit specification short-circuits yum's attempt to install BEA for Tomcat)
    tomcat5, tomcat5-webapps, tomcat5-admin-webapps (tomcat5-admin-webapps is probably not actually required, but is very handy)
    • Set-up Notes
    The best "tutorial" is http://wiki.apache.org/nutch/NutchTutorial
    Create /etc/profile.d/java.sh for "export JAVA_HOME=/etc/alternatives/jre" (required for the crawler, architecture independent definition)
    In /opt (in this example, though anyplace will do) "tar xzf /path/to/tar/file/nutch-1.0.tar.gz"
    Set http.agent.name in /opt/nutch-1.0/conf/nutch-site.xml (e.g., FedoraProject)
    Create a flat file of starting URLs in /opt/nutch-1.0/urls/crawl-start.txt (it can be any file name in the directory)
    Edit /opt/nutch-1.0/conf/crawl-urlfilter.txt to set regular expressions which keep or discard links for processing (e.g., the domain of the servers crawled)
    If not running the crawler as root (Why would you?), create directories /opt/nutch-1.0/crawl and /opt/nutch-1.0/logs writable by whatever uid/gid used to crawl
    Note: The crawler also needs to be able to create temporary files in its working directory
    Run the crawler, putting the database in /opt/nutch-1.0/crawl
    Deploy the WAR file from /opt/nutch-1.0/nutch-1.0.war as /nutch in Tomcat manager
    Set searcher.dir in /var/lib/tomcat5/webapps/nutch/WEB-INF/classes/nutch-site.xml to /opt/nutch-1.0/crawl
  • Swish-e [10] - akistler examined
    • Description
    written in C
    Note: Swish++ is a rewrite in C++ (not evaluated here)
    • Evaluation
    Search engine with a built-in web crawler, a built-in file system crawler, and an interface for an external crawler. The distribution includes sample search pages which use the Perl API. There is also a C API. The index is not customizable, but does include a facility for including metawords (exact match) and the path in the index for each document. The documentation acknowledges that the software only supports ASCII, but some MBCS may also work.
    • Requirements
    buildrequires
    • gcc
    • make
    • libxml2-devel
    • zlib-devel
    requires
    • libxml2
    • zlib
    • perl-libwww-perl (for the built-in spider)
    • others as desired to index documents (pdf, etc.)
  • Xapian [11] - akistler examined
    • Description
    written in C++
    bindings to Python, Ruby, and Perl XS
    xapian-core, xapian-bindings, and perl-Search-Xapian in Fedora already; xapian-omega is not
    additional bindings to PHP, Java, and more (?)
    Omega provides a Xapian front-end for indexing (via script) and searching (command line or CGI)
    Omega provides glue scripts for ht://Dig, mbox files, and perl DBI
    Flax [12] is another search engine built on top of Xapian and CherryPy
    • Evaluation
    Xapian is a search engine library. Omega adds functionality on top of Xapian. The Xapian database is very flexible, supporting an entirely user-designed schema. Usage through Omega loses very little, if any, of that flexibility, however the supplied Omega CGI is extremely rudimentary. The supplied Omega CGI also requires the database to be named "default," although that can be changed. Database columns are of type field or index. Fields are stored verbatim (e.g., URL, date, MIME type, keywords). Indices are input as blocks of text or other content to be indexed, but not stored (e.g., the corpus of a file or web page). The Omega scriptindex utility can be combined with an external web crawler for HTML. Making Omega work with Apache requires relabeling /var/lib/omega as httpd_sys_content, or moving /var/lib/omega to /var/www/omega and using the default context there. In this evaluation, /var/lib/omega was moved to /var/www/omega. Xapian only works with UTF-8.
    • Requirements
    xapian-core buildrequires
    • gcc gcc-c++
    • make
    • zlib-devel
    xapian-bindings buildrequires (not including gcc gcc-c++ make)
    • python python-devel
    • ruby ruby-devel
    • xapian-core-devel
    perl-Search-Xapian buildrequires (not including gcc gcc-c++ make)
    • perl
    • xapian-core-devel
    xapian-omega buildrequires (not including gcc gcc-c++ make)
    • libtool
    • xapian-core-devel
    xapian-core requires
    • xapian-core-libs
    xapian-bindings requires
    • coreutils
    • python
    • xapian-core-libs
    perl-Search-Xapian requries
    • perl
    • xapian-core-libs
    xapian-omega requires
    • httpd
    • perl
    • perl-DBI
    • xapian-core-libs
    • Set-up Notes
    Using custom crawler that requires perl-libwww-perl
    Updated Omega rpm .spec to move /var/lib/omega to /var/www/omega, including updating /etc/omega.conf
    Use default database at /var/www/omega/data/default
    /var/www/omega/data/default.index is
    URI : field boolean=Q unique=Q
    ContentType : field=MIMEtype index=T
    LastModified : field=Date date=unix
    Title : field index=S
    Content : index
    To index, run (as someone who can write to /var/www/omega/data/default)
    ./crawl.pl http://fedoraproject.org | scriptindex /var/www/omega/data/default /var/www/omega/data/default.index
    Note that "Content : unhtml index" would be preferable in the index, but unhtml apparently has bugs

Eliminated from Consideration

  • Reason for elimination
It has no crawling/spidering facility. It is a library toolkit (API) only.
  • Description
C++ port of Lucene
in Fedora already
described as beta by the developers
  • Isearch [14]
    • Reason for elimination
    It has no crawling/spidering facility. It is intended for local file indexing only.
    The web site appears to have fallen out of maintenance.
    • Description
    written in C++
  • Lucene [15] - akistler examined
  • Reason for elimination
It has no crawling/spidering facility. It has no user query interface. There are no samples.
  • Description
written in Java, but ported to others [16]
Requires/uses GCJ
in Fedora already
PyLucene [17] is a Python wrapper around Java Lucene
  • Evaluation
Search engine library meant to be integrated into applications
  • Requirements
buildrequires (based on 1.4.3-f7)
  • ant
  • ant-junit
  • java-1.4.2-gcj-compat-devel
  • javacc
  • jpackage-utils
  • junit
  • make
requires (based on 1.4.3-f7)
  • java-1.4.2-gcj-compat
  • Reason for elimination
It has no crawling/spidering facility. It indexes local documents only.
  • Description
written in Perl
in Fedora already
  • Solr [19] - akistler examined
  • Reason for elimination
It has no crawling/spidering facility. It has no user query interface. There are no samples.
  • Description
written in Java
based on Lucene
  • Evaluation
The documentation describes installing Sun Java to run Solr, but OpenJDK 1.5 or later is fine. Solr needs a Java servlet container in which to run. It comes with Jetty, but other containers should work, as well (e.g., Tomcat). Currently only supports UTF-8 characters.
Basically Solr provides an HTTP admin GUI for a search engine that uses a superset of the Lucene query syntax. The schema is very flexible. Set-up is essentially entirely through XML files. Applications can query the servlet port and get XML or JSON responses.
  • Requirements
buildrequires
  • ant (note that ant currently pulls in java-gcj-compat, too, but it appears not to be a problem)
  • ant-junit
  • java-1.6.0-openjdk-devel
  • junit
requires
  • java-1.6.0-openjdk
  • Terrier (TERabyte RetrIEveR) [20]
  • Reason for elimination
It has no crawler or user search tool. It does not run as a service (as provided), only interactively.
  • Description
written in Java
runs from the command line (i.e., not a Tomcat servlet)
  • YaCy [21] - huzaifa examined
  • Reason for elimination
Requires Sun Java
  • written in Java
  • well maintained
  • support for peer search engine database exchanges
  • customized search parameters
  • fast indexing and web interface for querying the back-end db
  • Reason for elimination
No crawling capability, only indexes local documents
User search/retrieval tool is command-line only, no web interface
  • Description
written in C

Never Considered

  • written in Ruby
  • specializes in source code search
  • not actively maintained
  • written in C#
  • written in Java
  • archives content rather than indexing it
  • written in C++
  • not actively maintained
  • It's just a MediaWiki plugin, not suitable for searching non-wiki sites
  • Requires EzMwLucene (Java) to be running on the servers to be searched
  • EzMwLucene is wiki-only, therefore MWSearch is wiki-only
  • written in Perl or TCL on top of PostgreSQL
  • Python interface available
  • not actively maintained
  • Perl port of Lucene
  • not actively maintained
  • Crawls the MediaWiki database, not the web site
  • Doesn't work for non-MediaWiki web sites, including any non-wiki web site
  • written in C++
  • designed to index SQL tables, not web pages
  • Written in C++
  • MediaWiki plug-in, so it's wiki-only
  • written in Python
  • inspired by Lucene, but closer to a Python port of parts of KinoSearch combined with some features of Terrier
  • toolkit only, not even sample crawlers and user interfaces

Public Testing

Public Testing is taking place on publictest3.

Search Engines

  • DataparkSearch (http://publictest3.fedoraproject.org/cgi-bin/dpsearch)
PostgreSQL installed (postgresql, postgresql-devel, postgresql-libs, postgresql-server)
SELinux (not present on publictest3, but needed eventually) needs:
"setsebool -P httpd_can_network_connect=1" to connect to PostgreSQL
Crawling trials (with database cleared each time, i.e., not incremental)
Memory and CPU utilization are modest, less than 10% each. Most CPU time is spent in I/O Wait for the database.
Depth=4, 2.5 hrs crawling, 2k documents, db = 320M (700M with clone detection off)
Depth=5, 16.5 hrs crawling, 12k documents, db = 1.1G
  • mnoGoSearch (http://publictest3.fedoraproject.org/cgi-bin/mnoGoSearch)
PostgreSQL installed (postgresql, postgresql-devel, postgresql-libs, postgresql-server)
SELinux (not present on publictest3, but needed eventually) needs:
"setsebool -P httpd_can_network_connect=1" to connect to PostgreSQL
Crawling trials (with database cleared each time, i.e., not incremental)
Memory and CPU utilization are quite modest, about 1% each. Most CPU time is spent in I/O Wait for the database.
Depth=4, 2 hrs crawling, 1.5 min indexing, 11k documents
Depth=5, 4.5 hrs crawling, 12 min indexing, 25k documents
Depth=6, 6.5 hrs crawling, 16 min indexing, 34k documents
Depth=7, 12 hrs crawling, 23 min indexing, 40k documents, db = 2.6G
  • Nutch (http://publictest3.fedoraproject.org/nutch)
The Nutch tarball was unpacked in /opt/nutch-1.0, just as in preliminary local testing
Tomcat is reverse proxied through Apache (see notes below)
Nutch's definition/conception of depth appears to be unusual. The crawler must be directed to spider much more deeply than should be necessary.
Crawls are executed as (e.g.) "/opt/nutch-1.0/bin/nutch crawl /opt/nutch-1.0/urls -dir /opt/nutch-1.0/crawl -depth 5 -threads 2"
Crawling trials
java process uses about 18% of 6G of memory (4G RAM, 2G swap), regardless of depth
Depth=4, 2 threads, 1.5k documents
Depth=5, 2 threads, 3 hrs, 8k documents
Depth=6, 1 thread, 8.5 hrs, 23k documents
Depth=7, 1 thread, 14.5 hrs, 37k documents, db = 400M
Depth=8, 1 thread, 16.5 hrs, 44k documents, db = 440M
  • Xapian (http://publictest3.fedoraproject.org/cgi-bin/omega)
At this time, only installed xapian-core-libs, xapian-core, and xapian-omega (i.e., no xapian-bindings or perl-Search-xapian)
Enabled cgi-bin in /etc/httpd/conf.d/cgi-bin.conf (see notes below)
Omega bombs on http://fedoraproject.org/wiki/Overview (and would possibly on others later) with "unhtml index"
Resolution: Don't use "unhtml"
Omega bombs on long URIs (longer than 244 chars)
Example: http://fedoraproject.org/wiki/Special:WhatLinksHere/Ru_RU/%D0%9F%D0%BB%D0%B0%D0%BD_%D1%80%D0%B0%D0%B1%D0%BE%D1%82%D1%8B_%D0%BF%D0%BE_%D0%BF%D0%B5%D1%80%D0%B5%D0%B2%D0%BE%D0%B4%D1%83_web-%D1%81%D0%B0%D0%B9%D1%82%D0%B0_%D0%BF%D1%80%D0%BE%D0%B5%D0%BA%D1%82%D0%B0_Fedora
Resolution: Enhanced custom crawler to filter URIs better (fedoraproject.org/w/ from the wiki); added capability to discard URIs that are too long (mostly hex URIs translated from other DBCS)
Perl custom crawler prints warnings for (and refuses to translate) URIs with Unicode characters outside the Latin 1 range
Resolution: None. This issue is known for URI.pm.[35]
Crawling trials
Failed, Depth=5, scriptindex used 70% of 4G of memory (2G RAM, 2G swap, 0% free)
terminated, system sluggish with swap I/O
Depth=4, 4 hrs, 15218 documents, index is about 500M on disk, scriptindex used 40% of 4G of memory (2G RAM, 2G swap)
Depth=5, 8 hrs, 41171 documents, index is about 1G on disk, scriptindex used 20% of 6G of memory (4G RAM, 2G swap)

Apache Configuration Notes

  • CGI for Xapian Omega and mnoGoSearch
In /etc/httpd/conf.d/cgi-bin.conf, just use the default configuration normally commented out in httpd.conf.
 # ScriptAlias: This controls which directories contain server scripts.
 # ScriptAliases are essentially the same as Aliases, except that
 # documents in the realname directory are treated as applications and
 # run by the server when requested rather than as documents sent to the client.
 # The same rules about trailing "/" apply to ScriptAlias directives as to
 # Alias.
 #
 ScriptAlias /cgi-bin/ "/var/www/cgi-bin/"
 #
 # "/var/www/cgi-bin" should be changed to whatever your ScriptAliased
 # CGI directory exists, if you have that configured.
 #
 <Directory "/var/www/cgi-bin">
     AllowOverride None
     Options None
     Order allow,deny
     Allow from all
 </Directory>
  • Reverse proxy for Tomcat
In /etc/httpd/conf.d/tomcat5.conf:
 <Location /admin>
   Order Allow,Deny
   Allow from ...
 </Location>
 ProxyPass        /admin   http://localhost:8082/admin
 ProxyPassReverse /admin   http://localhost:8082/admin
 #
 <Location /manager>
   Order Allow,Deny
   Allow from ...
 </Location>
 ProxyPass        /manager http://localhost:8082/manager
 ProxyPassReverse /manager http://localhost:8082/manager
 #
 ProxyPass        /nutch   http://localhost:8082/nutch
 ProxyPassReverse /nutch   http://localhost:8082/nutch

Tomcat Configuration Notes

  • In /etc/tomcat5/server.xml:
comment out the AJP connector on port 8009
comment out the HTTP connector on port 8080
uncomment the proxied HTTP connector on 8082
add proxyName to the HTTP connector on 8082
could alternatively define proxyName and proxyPort and undefine redirectPort in the HTTP connector on port 8080
  • SELinux (not present on publictest3, but needed eventually)
For port 8082, SELinux needs "setsebool -P httpd_can_network_connect=1"
Alternatively, for port 8080, SELinux needs "setsebool -P httpd_can_network_relay=1"
  • manager and admin (in fact, all users) defined in /etc/tomcat5/tomcat-users.xml
  • Recommended, but not done here, change the shutdown password in server.xml (default is SHUTDOWN)

Deployment Plan

<tbd>

References

  1. "Fedora Search Engine". Infrastructure Trac. https://fedorahosted.org/fedora-infrastructure/ticket/1055. 
  2. "DataparkSearch". DataparkSearch. http://www.dataparksearch.org/. 
  3. "Egothor". Egothor. http://www.egothor.org/. 
  4. "Ferret". David Balmain. http://ferret.davebalmain.com/. 
  5. 5.0 5.1 "Lucy". Apache Software Foundation. http://lucene.apache.org/lucy/. 
  6. "Indri". The Lemur Project. http://www.lemurproject.org/indri/. 
  7. "KinoSearch". Rectangular Research. http://www.rectangular.com/kinosearch/. 
  8. "mnoGoSearch". LavTech. http://www.mnogosearch.org/. 
  9. "Nutch". Apache Software Foundation. http://lucene.apache.org/nutch/. 
  10. "Swish-e". Swish-e. http://swish-e.org/. 
  11. "Xapian". Xapian Project. http://xapian.org/. 
  12. "Flax". Flax. http://www.flax.co.uk/products.shtml. 
  13. "CLucene". CLucene Project. http://sourceforge.net/projects/clucene/. 
  14. "Isearch". Isite. http://isite.awcubed.com/. 
  15. "Lucene". Apache Software Foundation. http://lucene.apache.org/. 
  16. "Lucene Implementations". Apache Software Foundation. http://wiki.apache.org/lucene-java/LuceneImplementations. 
  17. "PyLucene". Apache Software Foundation. http://lucene.apache.org/pylucene/. 
  18. "Namazu". Namazu Project. http://www.namazu.org/. 
  19. "Solr". Apache Software Foundation. http://lucene.apache.org/solr/. 
  20. "Terrier". Terrier Project. http://ir.dcs.gla.ac.uk/terrier/. 
  21. "YaCy". Karlsruhe Institute of Technology. http://yacy.net/. 
  22. "Zettair". Search Engine Group, Royal Melbourne Institute of Technology. http://www.seg.rmit.edu.au/zettair/. 
  23. "Gonzui". SourceForge. http://gonzui.sourceforge.net/. 
  24. "Grub". Wikia, Inc. http://grub.org/. 
  25. "Heritrix". Internet Archive. http://crawler.archive.org/. 
  26. "ht://Dig". The ht://Dig Group. http://www.htdig.org/. 
  27. "HtdigSearch Extension". MediaWiki. https://secure.wikimedia.org/wikipedia/mediawiki/wiki/Extension:HtdigSearch. 
  28. "MWSearch Extension". MediaWiki. https://secure.wikimedia.org/wikipedia/mediawiki/wiki/Extension:MWSearch. 
  29. "OpenFTS". XWare. http://www.astronet.ru/xware/#fts. 
  30. "Plucene". CPAN. http://search.cpan.org/~tmtm/Plucene-1.25. 
  31. "RigorousSearch Extension". MediaWiki. https://secure.wikimedia.org/wikipedia/mediawiki/wiki/Extension:RigorousSearch. 
  32. "Sphinx". Sphinx Technologies. http://sphinxsearch.com/. 
  33. "SphinxSearch Extension". MediaWiki. https://secure.wikimedia.org/wikipedia/mediawiki/wiki/Extension:SphinxSearch. 
  34. "Whoosh". Matt Chaput. http://whoosh.ca/. 
  35. ""URI.pm error"". Usenet. http://www.nntp.perl.org/group/perl.libwww/2006/01/msg6540.html.