LP 1827051: Remove Java Code

[Evergreen.git] / docs / modules / development / pages / intro_opensrf.adoc

= Easing gently into OpenSRF =
:toc:

== Abstract ==
The Evergreen open-source library system serves library consortia composed of
hundreds of branches with millions of patrons - for example,
http://www.georgialibraries.org/statelibrarian/bythenumbers.pdf[the Georgia
Public Library Service PINES system]. One of the claimed advantages of
Evergreen over alternative integrated library systems is the underlying Open
Service Request Framework (OpenSRF, pronounced "open surf") architecture. This
article introduces OpenSRF, demonstrates how to build OpenSRF services through
simple code examples, and explains the technical foundations on which OpenSRF
is built.

== Introducing OpenSRF ==
OpenSRF is a message routing network that offers scalability and failover
support for individual services and entire servers with minimal development and
deployment overhead. You can use OpenSRF to build loosely-coupled applications
that can be deployed on a single server or on clusters of geographically
distributed servers using the same code and minimal configuration changes.
Although copyright statements on some of the OpenSRF code date back to Mike
Rylander's original explorations in 2000, Evergreen was the first major
application to be developed with, and to take full advantage of, the OpenSRF
architecture starting in 2004. The first official release of OpenSRF was 0.1 in
February 2005 (http://evergreen-ils.org/blog/?p=21), but OpenSRF's development
continues a steady pace of enhancement and refinement, with the release of
1.0.0 in October 2008 and the most recent release of 1.2.2 in February 2010.

OpenSRF is a distinct break from the architectural approach used by previous
library systems and has more in common with modern Web applications. The
traditional "scale-up" approach to serve more transactions is to purchase a
server with more CPUs and more RAM, possibly splitting the load between a Web
server, a database server, and a business logic server. Evergreen, however, is
built on the Open Service Request Framework (OpenSRF) architecture, which
firmly embraces the "scale-out" approach of spreading transaction load over
cheap commodity servers. The http://evergreen-ils.org/blog/?p=56[initial GPLS
PINES hardware cluster], while certainly impressive, may have offered the
misleading impression that Evergreen is complex and requires a lot of hardware
to run. 

This article hopes to correct any such lingering impression by demonstrating
that OpenSRF itself is an extremely simple architecture on which one can easily
build applications of many kinds – not just library applications – and that you
can use a number of different languages to call and implement OpenSRF methods
with a minimal learning curve. With an application built on OpenSRF, when you
identify a bottleneck in your application's business logic layer, you can
adjust the number of the processes serving that particular bottleneck on each
of your servers; or if the problem is that your service is resource-hungry, you
could add an inexpensive server to your cluster and dedicate it to running that
resource-hungry service.

=== Programming language support ===

If you need to develop an entirely new OpenSRF service, you can choose from a
number of different languages in which to implement that service. OpenSRF
client language bindings have been written for C, Java, JavaScript, Perl, and
Python, and server language bindings have been written for C, Perl, and Python.
This article uses Perl examples as a lowest common denominator programming
language. Writing an OpenSRF binding for another language is a relatively small
task if that language offers libraries that support the core technologies on
which OpenSRF depends:

  * http://tools.ietf.org/html/rfc3920[Extensible Messaging and Presence
Protocol] (XMPP, sometimes referred to as Jabber) - provides the base messaging
infrastructure between OpenSRF clients and servers
  * http://json.org[JavaScript Object Notation] (JSON) - serializes the content
of each XMPP message in a standardized and concise format
  * http://memcached.org[memcached] - provides the caching service
  * http://tools.ietf.org/html/rfc5424[syslog] - the standard UNIX logging
service

Unfortunately, the
http://evergreen-ils.org/dokuwiki/doku.php?id=osrf-devel:primer[OpenSRF
reference documentation], although augmented by the
http://evergreen-ils.org/dokuwiki/doku.php?id=osrf-devel:primer[OpenSRF
glossary], blog posts like http://evergreen-ils.org/blog/?p=36[the description
of OpenSRF and Jabber], and even this article, is not a sufficient substitute
for a complete specification on which one could implement a language binding.
The recommended option for would-be developers of another language binding is
to use the Python implementation as the cleanest basis for a port to another
language.

=== OpenSRF communication flows over XMPP ===

The XMPP messaging service underpins OpenSRF, requiring an XMPP server such
as http://www.ejabberd.im/[ejabberd].  When you start OpenSRF, the first XMPP
clients that connect to the XMPP server are the OpenSRF public and private
_routers_. OpenSRF routers maintain a list of available services and connect
clients to available services. When an OpenSRF service starts, it establishes a
connection to the XMPP server and registers itself with the private router. The
OpenSRF configuration contains a list of public OpenSRF services, each of which
must also register with the public router. Services and clients connect to the
XMPP server using a single set of XMPP client credentials (for example,
`opensrf@private.localhost`), but use XMPP resource identifiers to
differentiate themselves in the Jabber ID (JID) for each connection. For
example, the JID for a copy of the `opensrf.simple-text` service with process
ID `6285` that has connected to the `private.localhost` domain using the
`opensrf` XMPP client credentials could be
`opensrf@private.localhost/opensrf.simple-text_drone_at_localhost_6285`.

[#OpenSRFOverHTTP]
=== OpenSRF communication flows over HTTP ===
Any OpenSRF service registered with the public router is accessible via the
OpenSRF HTTP Translator. The OpenSRF HTTP Translator implements the
http://www.open-ils.org/dokuwiki/doku.php?id=opensrf_over_http[OpenSRF-over-HTTP
proposed specification] as an Apache module that translates HTTP requests into
OpenSRF requests and returns OpenSRF results as HTTP results to the initiating
HTTP client. 

.Issuing an HTTP POST request to an OpenSRF method via the OpenSRF HTTP Translator
[source,bash]
--------------------------------------------------------------------------------
# curl request broken up over multiple lines for legibility
curl -H "X-OpenSRF-service: opensrf.simple-text"                                \ # <1>
    --data 'osrf-msg=[                                                          \ # <2>
        {"__c":"osrfMessage","__p":{"threadTrace":0,"locale":"en-CA",           \ # <3>
            "type":"REQUEST","payload": {"__c":"osrfMethod","__p":              \
                {"method":"opensrf.simple-text.reverse","params":["foobar"]}    \
            }}                                                                  \
        }]'                                                                     \
http://localhost/osrf-http-translator                                           \ # <4>
--------------------------------------------------------------------------------

<1> The `X-OpenSRF-service` header identifies the OpenSRF service of interest.

<2> The POST request consists of a single parameter, the `osrf-msg` value,
which contains a JSON array.

<3> The first object is an OpenSRF message (`"__c":"osrfMessage"`) with a set of
parameters (`"__p":{}`) containing:

  * the identifier for the request (`"threadTrace":0`); this value is echoed
back in the result

  * the message type (`"type":"REQUEST"`) 

  * the locale for the message; if the OpenSRF method is locale-sensitive, it
can check the locale for each OpenSRF request and return different information
depending on the locale

  * the payload of the message (`"payload":{}`) containing the OpenSRF method
request (`"__c":"osrfMethod"`) and its parameters (`"__p:"{}`), which in turn
contains:

  ** the method name for the request (`"method":"opensrf.simple-text.reverse"`)

  ** a set of JSON parameters to pass to the method (`"params":["foobar"]`); in
this case, a single string `"foobar"`

<4> The URL on which the OpenSRF HTTP translator is listening,
`/osrf-http-translator` is the default location in the Apache example
configuration files shipped with the OpenSRF source, but this is configurable.

[#httpResults]
.Results from an HTTP POST request to an OpenSRF method via the OpenSRF HTTP Translator
[source,bash]
--------------------------------------------------------------------------------
# HTTP response broken up over multiple lines for legibility
[{"__c":"osrfMessage","__p":                                                    \ # <1>
    {"threadTrace":0, "payload":                                                \ # <2> 
        {"__c":"osrfResult","__p":                                              \ # <3>
            {"status":"OK","content":"raboof","statusCode":200}                 \ # <4>
        },"type":"RESULT","locale":"en-CA"                                      \ # <5>
    }
},
{"__c":"osrfMessage","__p":                                                     \ # <6>
    {"threadTrace":0,"payload":                                                 \ # <7>
        {"__c":"osrfConnectStatus","__p":                                       \ # <8>
            {"status":"Request Complete","statusCode":205}                      \ # <9>
        },"type":"STATUS","locale":"en-CA"                                      \ # <10>
    }
}]
--------------------------------------------------------------------------------

<1> The OpenSRF HTTP Translator returns an array of JSON objects in its
response. Each object in the response is an OpenSRF message
(`"__c":"osrfMessage"`) with a collection of response parameters (`"__p":`).

<2> The OpenSRF message identifier (`"threadTrace":0`) confirms that this
message is in response to the request matching the same identifier.

<3> The message includes a payload JSON object (`"payload":`) with an OpenSRF
result for the request (`"__c":"osrfResult"`).

<4> The result includes a status indicator string (`"status":"OK"`), the content
of the result response - in this case, a single string "raboof"
(`"content":"raboof"`) - and an integer status code for the request
(`"statusCode":200`).

<5> The message also includes the message type (`"type":"RESULT"`) and the
message locale (`"locale":"en-CA"`).

<6> The second message in the set of results from the response.

<7> Again, the message identifier confirms that this message is in response to
a particular request.

<8> The payload of the message denotes that this message is an
OpenSRF connection status message (`"__c":"osrfConnectStatus"`), with some
information about the particular OpenSRF connection that was used for this
request.

<9> The response parameters for an OpenSRF connection status message include a
verbose status (`"status":"Request Complete"`) and an integer status code for
the connection status (`"statusCode":205).

<10> The message also includes the message type (`"type":"RESULT"`) and the
message locale (`"locale":"en-CA"`).


[TIP]
Before adding a new public OpenSRF service, ensure that it does
not introduce privilege escalation or unchecked access to data. For example,
the Evergreen `open-ils.cstore` private service is an object-relational mapper
that provides read and write access to the entire Evergreen database, so it
would be catastrophic to expose that service publicly. In comparison, the
Evergreen `open-ils.pcrud` public service offers the same functionality as
`open-ils.cstore` to any connected HTTP client or OpenSRF client, but the
additional authentication and authorization layer in `open-ils.pcrud` prevents
unchecked access to Evergreen's data.

=== Stateless and stateful connections ===

OpenSRF supports both _stateless_ and _stateful_ connections.  When an OpenSRF
client issues a `REQUEST` message in a _stateless_ connection, the router
forwards the request to the next available service and the service returns the
result directly to the client. 

.REQUEST flow in a stateless connection
image:media/REQUEST.png[REQUEST flow in a stateless connection]

When an OpenSRF client issues a `CONNECT` message to create a _stateful_ connection, the
router returns the Jabber ID of the next available service to the client so
that the client can issue one or more `REQUEST` message directly to that
particular service and the service will return corresponding `RESULT` messages
directly to the client. Until the client issues a `DISCONNECT` message, that
particular service is only available to the requesting client. Stateful connections
are useful for clients that need to make many requests from a particular service,
as it avoids the intermediary step of contacting the router for each request, as
well as for operations that require a controlled sequence of commands, such as a
set of database INSERT, UPDATE, and DELETE statements within a transaction.

.CONNECT, REQUEST, and DISCONNECT flow in a stateful connection
image:media/CONNECT.png[CONNECT, REQUEST, and DISCONNECT flow in a stateful connection]

== Enough jibber-jabber: writing an OpenSRF service ==
Imagine an application architecture in which 10 lines of Perl or Python, using
the data types native to each language, are enough to implement a method that
can then be deployed and invoked seamlessly across hundreds of servers.  You
have just imagined developing with OpenSRF – it is truly that simple. Under the
covers, of course, the OpenSRF language bindings do an incredible amount of
work on behalf of the developer. An OpenSRF application consists of one or more
OpenSRF services that expose methods: for example, the `opensrf.simple-text`
http://git.evergreen-ils.org/?p=OpenSRF.git;a=blob_plain;f=src/perl/lib/OpenSRF/Application/Demo/SimpleText.pm[demonstration
service] exposes the `opensrf.simple-text.split()` and
`opensrf.simple-text.reverse()` methods. Each method accepts zero or more
arguments and returns zero or one results. The data types supported by OpenSRF
arguments and results are typical core language data types: strings, numbers,
booleans, arrays, and hashes.

To implement a new OpenSRF service, perform the following steps:

  1. Include the base OpenSRF support libraries
  2. Write the code for each of your OpenSRF methods as separate procedures 
  3. Register each method
  4. Add the service definition to the OpenSRF configuration files

For example, the following code implements an OpenSRF service. The service
includes one method named `opensrf.simple-text.reverse()` that accepts one
string as input and returns the reversed version of that string:

[source,perl]
--------------------------------------------------------------------------------
#!/usr/bin/perl

package OpenSRF::Application::Demo::SimpleText;

use strict;

use OpenSRF::Application;
use parent qw/OpenSRF::Application/;

sub text_reverse {
    my ($self , $conn, $text) = @_;
    my $reversed_text = scalar reverse($text);
    return $reversed_text;
}

__PACKAGE__->register_method(
    method    => 'text_reverse',
    api_name  => 'opensrf.simple-text.reverse'
);
--------------------------------------------------------------------------------

Ten lines of code, and we have a complete OpenSRF service that exposes a single
method and could be deployed quickly on a cluster of servers to meet your
application's ravenous demand for reversed strings! If you're unfamiliar with
Perl, the `use OpenSRF::Application; use parent qw/OpenSRF::Application/;`
lines tell this package to inherit methods and properties from the
`OpenSRF::Application` module. For example, the call to
`__PACKAGE__->register_method()` is defined in `OpenSRF::Application` but due to
inheritance is available in this package (named by the special Perl symbol
`__PACKAGE__` that contains the current package name). The `register_method()`
procedure is how we introduce a method to the rest of the OpenSRF world.

[#serviceRegistration]
=== Registering a service with the OpenSRF configuration files ===

Two files control most of the configuration for OpenSRF:

  * `opensrf.xml` contains the configuration for the service itself as well as
a list of which application servers in your OpenSRF cluster should start
the service
  * `opensrf_core.xml` (often referred to as the "bootstrap configuration"
file) contains the OpenSRF networking information, including the XMPP server
connection credentials for the public and private routers; you only need to touch
this for a new service if the new service needs to be accessible via the
public router

Begin by defining the service itself in `opensrf.xml`. To register the
`opensrf.simple-text` service, add the following section to the `<apps>`
element (corresponding to the XPath `/opensrf/default/apps/`):

[source,xml]
--------------------------------------------------------------------------------
<apps>
  <opensrf.simple-text>                                                      <!-- <1> -->
    <keepalive>3</keepalive>                                                 <!-- <2> -->
    <stateless>1</stateless>                                                 <!-- <3> -->        
    <language>perl</language>                                                <!-- <4> -->
    <implementation>OpenSRF::Application::Demo::SimpleText</implementation>  <!-- <5> -->
    <max_requests>100</max_requests>                                         <!-- <6> -->
    <unix_config>
      <max_requests>1000</max_requests>                                      <!-- <7> -->
      <unix_log>opensrf.simple-text_unix.log</unix_log>                      <!-- <8> -->
      <unix_sock>opensrf.simple-text_unix.sock</unix_sock>                   <!-- <9> -->
      <unix_pid>opensrf.simple-text_unix.pid</unix_pid>                     <!-- <10> -->
      <min_children>5</min_children>                                        <!-- <11> -->
      <max_children>15</max_children>                                       <!-- <12> -->
      <min_spare_children>2</min_spare_children>                            <!-- <13> -->
      <max_spare_children>5</max_spare_children>                            <!-- <14> -->
    </unix_config>
  </opensrf.simple-text>

  <!-- other OpenSRF services registered here... -->
</apps>
--------------------------------------------------------------------------------

<1> The element name is the name that the OpenSRF control scripts use to refer
to the service.

<2> Specifies the interval (in seconds) between checks to determine if the
service is still running.

<3> Specifies whether OpenSRF clients can call methods from this service
without first having to create a connection to a specific service backend
process for that service. If the value is `1`, then the client can simply
issue a request and the router will forward the request to an available
service and the result will be returned directly to the client.

<4> Specifies the programming language in which the service is implemented

<5> Specifies the name of the library or module in which the service is implemented

<6> (C implementations): Specifies the maximum number of requests a process
serves before it is killed and replaced by a new process.

<7> (Perl implementations): Specifies the maximum number of requests a process
serves before it is killed and replaced by a new process.

<8> The name of the log file for language-specific log messages such as syntax
warnings.

<9> The name of the UNIX socket used for inter-process communications.

<10> The name of the PID file for the master process for the service.

<11> The minimum number of child processes that should be running at any given
time.

<12> The maximum number of child processes that should be running at any given
time.

<13> The minimum number of child processes that should be available to handle
incoming requests. If there are fewer than this number of spare child
processes, new processes will be spawned.

<14>  The maximum number of child processes that should be available to handle
incoming requests. If there are more than this number of spare child processes,
the extra processes will be killed.

To make the service accessible via the public router, you must also
edit the `opensrf_core.xml` configuration file to add the service to the list
of publicly accessible services:

.Making a service publicly accessible in `opensrf_core.xml`
[source,xml]
--------------------------------------------------------------------------------
<router>                                                                     <!-- <1> --> 
    <!-- This is the public router. On this router, we only register applications
     which should be accessible to everyone on the opensrf network -->
    <name>router</name>
    <domain>public.localhost</domain>                                        <!-- <2> -->
    <services>
        <service>opensrf.math</service> 
        <service>opensrf.simple-text</service>                               <!-- <3> -->
    </services>
</router>
--------------------------------------------------------------------------------

<1> This section of the `opensrf_core.xml` file is located at XPath 
`/config/opensrf/routers/`.

<2> `public.localhost` is the canonical public router domain in the OpenSRF
installation instructions.

<3> Each `<service>` element contained in the `<services>` element 
offers their services via the public router as well as the private router.

Once you have defined the new service, you must restart the OpenSRF Router
to retrieve the new configuration and start or restart the service itself.

=== Calling an OpenSRF method ===
OpenSRF clients in any supported language can invoke OpenSRF services in any
supported language. So let's see a few examples of how we can call our fancy
new `opensrf.simple-text.reverse()` method:

==== Calling OpenSRF methods from the srfsh client ====
`srfsh` is a command-line tool installed with OpenSRF that you can use to call
OpenSRF methods. To call an OpenSRF method, issue the `request` command and pass
the OpenSRF service and method name as the first two arguments; then pass a list
of JSON objects as the arguments to the method being invoked.

The following example calls the `opensrf.simple-text.reverse` method of the
`opensrf.simple-text` OpenSRF service, passing the string `"foobar"` as the
only method argument:

[source,sh]
--------------------------------------------------------------------------------
$ srfsh
srfsh # request opensrf.simple-text opensrf.simple-text.reverse "foobar"

Received Data: "raboof"

=------------------------------------
Request Completed Successfully
Request Time in seconds: 0.016718
=------------------------------------
--------------------------------------------------------------------------------

[#opensrfIntrospection]
==== Getting documentation for OpenSRF methods from the srfsh client ====

The `srfsh` client also gives you command-line access to retrieving metadata
about OpenSRF services and methods. For a given OpenSRF method, for example,
you can retrieve information such as the minimum number of required arguments,
the data type and a description of each argument, the package or library in
which the method is implemented, and a description of the method. To retrieve
the documentation for an opensrf method from `srfsh`, issue the `introspect`
command, followed by the name of the OpenSRF service and (optionally) the
name of the OpenSRF method. If you do not pass a method name to the `introspect`
command, `srfsh` lists all of the methods offered by the service. If you pass
a partial method name, `srfsh` lists all of the methods that match that portion
of the method name.

[NOTE]
The quality and availability of the descriptive information for each
method depends on the developer to register the method with complete and
accurate information. The quality varies across the set of OpenSRF and
Evergreen APIs, although some effort is being put towards improving the
state of the internal documentation.

[source,sh]
--------------------------------------------------------------------------------
srfsh# introspect opensrf.simple-text "opensrf.simple-text.reverse"
--> opensrf.simple-text

Received Data: {
  "__c":"opensrf.simple-text",
  "__p":{
    "api_level":1,
    "stream":0,                                                      \ # <1>
    "object_hint":"OpenSRF_Application_Demo_SimpleText",
    "remote":0,
    "package":"OpenSRF::Application::Demo::SimpleText",              \ # <2>
    "api_name":"opensrf.simple-text.reverse",                        \ # <3>
    "server_class":"opensrf.simple-text",
    "signature":{                                                    \ # <4>
      "params":[                                                     \ # <5>
        {
          "desc":"The string to reverse",
          "name":"text",
          "type":"string"
        }
      ],
      "desc":"Returns the input string in reverse order\n",          \ # <6>
      "return":{                                                     \ # <7>
        "desc":"Returns the input string in reverse order",
        "type":"string"
      }
    },
    "method":"text_reverse",                                         \ # <8>
    "argc":1                                                         \ # <9>
  }
}
--------------------------------------------------------------------------------

<1> `stream` denotes whether the method supports streaming responses or not.

<2> `package` identifies which package or library implements the method.

<3> `api_name` identifies the name of the OpenSRF method.

<4> `signature` is a hash that describes the parameters for the method.

<5> `params` is an array of hashes describing each parameter in the method;
each parameter has a description (`desc`), name (`name`), and type (`type`).

<6> `desc` is a string that describes the method itself.

<7> `return` is a hash that describes the return value for the method; it
contains a description of the return value (`desc`) and the type of the
returned value (`type`).

<8> `method` identifies the name of the function or method in the source
implementation.

<9> `argc` is an integer describing the minimum number of arguments that
must be passed to this method.

==== Calling OpenSRF methods from Perl applications ====

To call an OpenSRF method from Perl, you must connect to the OpenSRF service,
issue the request to the method, and then retrieve the results.

[source,perl]
--------------------------------------------------------------------------------
#/usr/bin/perl
use strict;
use OpenSRF::AppSession;
use OpenSRF::System;

OpenSRF::System->bootstrap_client(config_file => '/openils/conf/opensrf_core.xml'); # <1>

my $session = OpenSRF::AppSession->create("opensrf.simple-text");                   # <2>

print "substring: Accepts a string and a number as input, returns a string\n";
my $result = $session->request("opensrf.simple-text.substring", "foobar", 3);       # <3>
my $request = $result->gather();                                                    # <4>
print "Substring: $request\n\n";

print "split: Accepts two strings as input, returns an array of strings\n";
$request = $session->request("opensrf.simple-text.split", "This is a test", " ");   # <5>
my $output = "Split: [";
my $element;
while ($element = $request->recv()) {                                               # <6>
    $output .= $element->content . ", ";                                            # <7>
}
$output =~ s/, $/]/;
print $output . "\n\n";

print "statistics: Accepts an array of strings as input, returns a hash\n";
my @many_strings = [
    "First I think I'll have breakfast",
    "Then I think that lunch would be nice",
    "And then seventy desserts to finish off the day"
];

$result = $session->request("opensrf.simple-text.statistics", @many_strings);       # <8>
$request = $result->gather();                                                       # <9>
print "Length: " . $result->{'length'} . "\n";
print "Word count: " . $result->{'word_count'} . "\n";

$session->disconnect();                                                             # <10>
--------------------------------------------------------------------------------

<1> The `OpenSRF::System->bootstrap_client()` method reads the OpenSRF
configuration information from the indicated file and creates an XMPP client
connection based on that information.

<2> The `OpenSRF::AppSession->create()` method accepts one argument - the name
of the OpenSRF service to which you want to want to make one or more requests -
and returns an object prepared to use the client connection to make those
requests.

<3> The `OpenSRF::AppSession->request()` method accepts a minimum of one
argument - the name of the OpenSRF method to which you want to make a request -
followed by zero or more arguments to pass to the OpenSRF method as input
values. This example passes a string and an integer to the
`opensrf.simple-text.substring` method defined by the `opensrf.simple-text`
OpenSRF service.

<4> The `gather()` method, called on the result object returned by the
`request()` method, iterates over all of the possible results from the result
object and returns a single variable.

<5> This `request()` call passes two strings to the `opensrf.simple-text.split`
method defined by the `opensrf.simple-text` OpenSRF service and returns (via
`gather()`) a reference to an array of results.

<6> The `opensrf.simple-text.split()` method is a streaming method that
returns an array of results with one element per `recv()` call on the
result object. We could use the `gather()` method to retrieve all of the
results in a single array reference, but instead we simply iterate over
the result variable until there are no more results to retrieve.

<7> While the `gather()` convenience method returns only the content of the
complete set of results for a given request, the `recv()` method returns an
OpenSRF result object with `status`, `statusCode`, and `content` fields as
we saw in <<httpResults,the HTTP results example>>.

<8> This `request()` call passes an array to the
`opensrf.simple-text.statistics` method defined by the `opensrf.simple-text`
OpenSRF service.

<9> The result object returns a hash reference via `gather()`. The hash 
contains the `length` and `word_count` keys we defined in the method.

<10> The `OpenSRF::AppSession->disconnect()` method closes the XMPP client
connection and cleans up resources associated with the session.

=== Accepting and returning more interesting data types ===

Of course, the example of accepting a single string and returning a single
string is not very interesting. In real life, our applications tend to pass
around multiple arguments, including arrays and hashes. Fortunately, OpenSRF
makes that easy to deal with; in Perl, for example, returning a reference to
the data type does the right thing. In the following example of a method that
returns a list, we accept two arguments of type string: the string to be split,
and the delimiter that should be used to split the string.

.Text splitting method - streaming mode
[source,perl]
--------------------------------------------------------------------------------
sub text_split {
    my $self = shift;
    my $conn = shift;
    my $text = shift;
    my $delimiter = shift || ' ';

    my @split_text = split $delimiter, $text;
    return \@split_text;
}

__PACKAGE__->register_method(
    method    => 'text_split',
    api_name  => 'opensrf.simple-text.split'
);
--------------------------------------------------------------------------------

We simply return a reference to the list, and OpenSRF does the rest of the work
for us to convert the data into the language-independent format that is then
returned to the caller. As a caller of a given method, you must rely on the
documentation used to register to determine the data structures - if the developer has
added the appropriate documentation.

=== Accepting and returning Evergreen objects ===

OpenSRF is agnostic about objects; its role is to pass JSON back and forth
between OpenSRF clients and services, and it allows the specific clients and
services to define their own semantics for the JSON structures. On top of that
infrastructure, Evergreen offers the fieldmapper: an object-relational mapper
that provides a complete definition of all objects, their properties, their
relationships to other objects, the permissions required to create, read,
update, or delete objects of that type, and the database table or view on which
they are based.

The Evergreen fieldmapper offers a great deal of convenience for working with
complex system objects beyond the basic mapping of classes to database
schemas. Although the result is passed over the wire as a JSON object
containing the indicated fields, fieldmapper-aware clients then turn those
JSON objects into native objects with setter / getter methods for each field.

All of this metadata about Evergreen objects is defined in the
fieldmapper configuration file (`/openils/conf/fm_IDL.xml`), and access to
these classes is provided by the `open-ils.cstore`, `open-ils.pcrud`, and
`open-ils.reporter-store` OpenSRF services which parse the fieldmapper
configuration file and dynamically register OpenSRF methods for creating,
reading, updating, and deleting all of the defined classes.

.Example fieldmapper class definition for "Open User Summary"
[source,xml]
--------------------------------------------------------------------------------
<class id="mous" controller="open-ils.cstore open-ils.pcrud"
 oils_obj:fieldmapper="money::open_user_summary"
 oils_persist:tablename="money.open_usr_summary"
 reporter:label="Open User Summary">                                       <!-- <1> -->
    <fields oils_persist:primary="usr" oils_persist:sequence="">           <!-- <2> -->
        <field name="balance_owed" reporter:datatype="money" />            <!-- <3> -->
        <field name="total_owed" reporter:datatype="money" />
        <field name="total_paid" reporter:datatype="money" />
        <field name="usr" reporter:datatype="link"/>
    </fields>
    <links>
        <link field="usr" reltype="has_a" key="id" map="" class="au"/>     <!-- <4> -->
    </links>
    <permacrud xmlns="http://open-ils.org/spec/opensrf/IDL/permacrud/v1">  <!-- <5> -->
        <actions>
            <retrieve permission="VIEW_USER">                              <!-- <6> -->
                <context link="usr" field="home_ou"/>                      <!-- <7> -->
            </retrieve>
        </actions>
    </permacrud>
</class>
--------------------------------------------------------------------------------

<1> The `<class>` element defines the class:

  * The `id` attribute defines the _class hint_ that identifies the class both
elsewhere in the fieldmapper configuration file, such as in the value of the
`field` attribute of the `<link>` element, and in the JSON object itself when
it is instantiated. For example, an "Open User Summary" JSON object would have
the top level property of `"__c":"mous"`.

  * The `controller` attribute identifies the services that have direct access
to this class. If `open-ils.pcrud` is not listed, for example, then there is
no means to directly access members of this class through a public service.

  * The `oils_obj:fieldmapper` attribute defines the name of the Perl
fieldmapper class that will be dynamically generated to provide setter and
getter methods for instances of the class.

  * The `oils_persist:tablename` attribute identifies the schema name and table
name of the database table that stores the data that represents the instances
of this class. In this case, the schema is `money` and the table is
`open_usr_summary`.

  * The `reporter:label` attribute defines a human-readable name for the class
used in the reporting interface to identify the class. These names are defined
in English in the fieldmapper configuration file; however, they are extracted
so that they can be translated and served in the user's language of choice.

<2> The `<fields>` element lists all of the fields that belong to the object.

  * The `oils_persist:primary` attribute identifies the field that acts as the
primary key for the object; in this case, the field with the name `usr`.

  * The `oils_persist:sequence` attribute identifies the sequence object
(if any) in this database provides values for new instances of this class. In
this case, the primary key is defined by a field that is linked to a different
table, so no sequence is used to populate these instances.

<3> Each `<field>` element defines a single field with the following attributes:

  * The `name` attribute identifies the column name of the field in the
underlying database table as well as providing a name for the setter / getter
method that can be invoked in the JSON or native version of the object.

  * The `reporter:datatype` attribute defines how the reporter should treat
the contents of the field for the purposes of querying and display.

  * The `reporter:label` attribute can be used to provide a human-readable name
for each field; without it, the reporter falls back to the value of the `name`
attribute.

<4> The `<links>` element contains a set of zero or more `<link>` elements,
each of which defines a relationship between the class being described and
another class.

  * The `field` attribute identifies the field named in this class that links
to the external class.

  * The `reltype` attribute identifies the kind of relationship between the
classes; in the case of `has_a`, each value in the `usr` field is guaranteed
to have a corresponding value in the external class.

  * The `key` attribute identifies the name of the field in the external
class to which this field links.

  * The rarely-used `map` attribute identifies a second class to which
the external class links; it enables this field to define a direct
relationship to an external class with one degree of separation, to
avoid having to retrieve all of the linked members of an intermediate
class just to retrieve the instances from the actual desired target class.

  * The `class` attribute identifies the external class to which this field
links.

<5> The `<permacrud>` element defines the permissions that must have been
granted to a user to operate on instances of this class.

<6> The `<retrieve>` element is one of four possible children of the
`<actions>` element that define the permissions required for each action:
create, retrieve, update, and delete.

  * The `permission` attribute identifies the name of the permission that must
have been granted to the user to perform the action.

  * The `contextfield` attribute, if it exists, defines the field in this class
that identifies the library within the system for which the user must have
privileges to work. If a user has been granted a given permission, but has not been
granted privileges to work at a given library, they can not perform the action
at that library.

<7> The rarely-used `<context>` element identifies a linked field (`link`
attribute) in this class which links to an external class that holds the field
(`field` attribute) that identifies the library within the system for which the
user must have privileges to work.

When you retrieve an instance of a class, you can ask for the result to
_flesh_ some or all of the linked fields of that class, so that the linked
instances are returned embedded directly in your requested instance. In that
same request you can ask for the fleshed instances to in turn have their linked
fields fleshed. By bundling all of this into a single request and result
sequence, you can avoid the network overhead of requiring the client to request
the base object, then request each linked object in turn.

You can also iterate over a collection of instances and set the automatically
generated `isdeleted`, `isupdated`, or `isnew` properties to indicate that
the given instance has been deleted, updated, or created respectively.
Evergreen can then act in batch mode over the collection to perform the
requested actions on any of the instances that have been flagged for action.

=== Returning streaming results ===

In the previous implementation of the `opensrf.simple-text.split` method, we
returned a reference to the complete array of results. For small values being
delivered over the network, this is perfectly acceptable, but for large sets of
values this can pose a number of problems for the requesting client. Consider a
service that returns a set of bibliographic records in response to a query like
"all records edited in the past month"; if the underlying database is
relatively active, that could result in thousands of records being returned as
a single network request. The client would be forced to block until all of the
results are returned, likely resulting in a significant delay, and depending on
the implementation, correspondingly large amounts of memory might be consumed
as all of the results are read from the network in a single block.

OpenSRF offers a solution to this problem. If the method returns results that
can be divided into separate meaningful units, you can register the OpenSRF
method as a streaming method and enable the client to loop over the results one
unit at a time until the method returns no further results. In addition to
registering the method with the provided name, OpenSRF also registers an additional
method with `.atomic` appended to the method name. The `.atomic` variant gathers
all of the results into a single block to return to the client, giving the caller
the ability to choose either streaming or atomic results from a single method
definition.

In the following example, the text splitting method has been reimplemented to
support streaming; very few changes are required:

.Text splitting method - streaming mode
[source,perl]
--------------------------------------------------------------------------------
sub text_split {
    my $self = shift;
    my $conn = shift;
    my $text = shift;
    my $delimiter = shift || ' ';

    my @split_text = split $delimiter, $text;
    foreach my $string (@split_text) {           # <1>
        $conn->respond($string);
    }
    return undef;
}

__PACKAGE__->register_method(
    method    => 'text_split',
    api_name  => 'opensrf.simple-text.split',
    stream    => 1                              # <2>
);
--------------------------------------------------------------------------------

<1> Rather than returning a reference to the array, a streaming method loops
over the contents of the array and invokes the `respond()` method of the
connection object on each element of the array.

<2> Registering the method as a streaming method instructs OpenSRF to also
register an atomic variant (`opensrf.simple-text.split.atomic`).

=== Error! Warning! Info! Debug! ===
As hard as it may be to believe, it is true: applications sometimes do not
behave in the expected manner, particularly when they are still under
development. The server language bindings for OpenSRF include integrated
support for logging messages at the levels of ERROR, WARNING, INFO, DEBUG, and
the extremely verbose INTERNAL to either a local file or to a syslogger
service. The destination of the log files, and the level of verbosity to be
logged, is set in the `opensrf_core.xml` configuration file. To add logging to
our Perl example, we just have to add the `OpenSRF::Utils::Logger` package to our
list of used Perl modules, then invoke the logger at the desired logging level.

You can include many calls to the OpenSRF logger; only those that are higher
than your configured logging level will actually hit the log. The following
example exercises all of the available logging levels in OpenSRF:

[source,perl]
--------------------------------------------------------------------------------
use OpenSRF::Utils::Logger;
my $logger = OpenSRF::Utils::Logger;
# some code in some function
{
    $logger->error("Hmm, something bad DEFINITELY happened!");
    $logger->warn("Hmm, something bad might have happened.");
    $logger->info("Something happened.");
    $logger->debug("Something happened; here are some more details.");
    $logger->internal("Something happened; here are all the gory details.")
}
--------------------------------------------------------------------------------

If you call the mythical OpenSRF method containing the preceding OpenSRF logger
statements on a system running at the default logging level of INFO, you will
only see the INFO, WARN, and ERR messages, as follows: 

.Results of logging calls at the default level of INFO
--------------------------------------------------------------------------------
[2010-03-17 22:27:30] opensrf.simple-text [ERR :5681:SimpleText.pm:277:] Hmm, something bad DEFINITELY happened!
[2010-03-17 22:27:30] opensrf.simple-text [WARN:5681:SimpleText.pm:278:] Hmm, something bad might have happened.
[2010-03-17 22:27:30] opensrf.simple-text [INFO:5681:SimpleText.pm:279:] Something happened.
--------------------------------------------------------------------------------

If you then increase the the logging level to INTERNAL (5), the logs will
contain much more information, as follows:

.Results of logging calls at the default level of INTERNAL
--------------------------------------------------------------------------------
[2010-03-17 22:48:11] opensrf.simple-text [ERR :5934:SimpleText.pm:277:] Hmm, something bad DEFINITELY happened!
[2010-03-17 22:48:11] opensrf.simple-text [WARN:5934:SimpleText.pm:278:] Hmm, something bad might have happened.
[2010-03-17 22:48:11] opensrf.simple-text [INFO:5934:SimpleText.pm:279:] Something happened.
[2010-03-17 22:48:11] opensrf.simple-text [DEBG:5934:SimpleText.pm:280:] Something happened; here are some more details.
[2010-03-17 22:48:11] opensrf.simple-text [INTL:5934:SimpleText.pm:281:] Something happened; here are all the gory details.
[2010-03-17 22:48:11] opensrf.simple-text [ERR :5934:SimpleText.pm:283:] Resolver did not find a cache hit
[2010-03-17 22:48:21] opensrf.simple-text [INTL:5934:Cache.pm:125:] Stored opensrf.simple-text.test_cache.masaa => "here" in memcached server
[2010-03-17 22:48:21] opensrf.simple-text [DEBG:5934:Application.pm:579:] Coderef for [OpenSRF::Application::Demo::SimpleText::test_cache] has been run
[2010-03-17 22:48:21] opensrf.simple-text [DEBG:5934:Application.pm:586:] A top level Request object is responding de nada
[2010-03-17 22:48:21] opensrf.simple-text [DEBG:5934:Application.pm:190:] Method duration for [opensrf.simple-text.test_cache]:  10.005
[2010-03-17 22:48:21] opensrf.simple-text [INTL:5934:AppSession.pm:780:] Calling queue_wait(0)
[2010-03-17 22:48:21] opensrf.simple-text [INTL:5934:AppSession.pm:769:] Resending...0
[2010-03-17 22:48:21] opensrf.simple-text [INTL:5934:AppSession.pm:450:] In send
[2010-03-17 22:48:21] opensrf.simple-text [DEBG:5934:AppSession.pm:506:] AppSession sending RESULT to opensrf@private.localhost/_dan-karmic-liblap_1268880489.752154_5943 with threadTrace [1]
[2010-03-17 22:48:21] opensrf.simple-text [DEBG:5934:AppSession.pm:506:] AppSession sending STATUS to opensrf@private.localhost/_dan-karmic-liblap_1268880489.752154_5943 with threadTrace [1]
...
--------------------------------------------------------------------------------

To see everything that is happening in OpenSRF, try leaving your logging level
set to INTERNAL for a few minutes - just ensure that you have a lot of free disk
space available if you have a moderately busy system!

=== Caching results: one secret of scalability ===
If you have ever used an application that depends on a remote Web service
outside of your control-say, if you need to retrieve results from a
microblogging service-you know the pain of latency and dependability (or the
lack thereof). To improve response time in OpenSRF applications, you can take
advantage of the support offered by the `OpenSRF::Utils::Cache` module for
communicating with a local instance or cluster of memcache daemons to store
and retrieve persistent values.

[source,perl]
--------------------------------------------------------------------------------
use OpenSRF::Utils::Cache;                                       # <1>
sub test_cache {
    my $self = shift;
    my $conn = shift;
    my $test_key = shift;
    my $cache = OpenSRF::Utils::Cache->new('global');            # <2>
    my $cache_key = "opensrf.simple-text.test_cache.$test_key";  # <3>
    my $result = $cache->get_cache($cache_key) || undef;         # <4>
    if ($result) {
        $logger->info("Resolver found a cache hit");
        return $result;
    }
    sleep 10;                                                    # <5>
    my $cache_timeout = 300;                                     # <6>
    $cache->put_cache($cache_key, "here", $cache_timeout);       # <7>
    return "There was no cache hit.";
}
--------------------------------------------------------------------------------

This example:

<1> Imports the OpenSRF::Utils::Cache module

<2> Creates a cache object

<3> Creates a unique cache key based on the OpenSRF method name and
request input value

<4> Checks to see if the cache key already exists; if so, it immediately
returns that value

<5> If the cache key does not exist, the code sleeps for 10 seconds to
simulate a call to a slow remote Web service, or an intensive process

<6> Sets a value for the lifetime of the cache key in seconds

<7> When the code has retrieved its value, then it can create the cache
entry, with the cache key, value to be stored ("here"), and the timeout
value in seconds to ensure that we do not return stale data on subsequent
calls

=== Initializing the service and its children: child labour ===
When an OpenSRF service is started, it looks for a procedure called
`initialize()` to set up any global variables shared by all of the children of
the service. The `initialize()` procedure is typically used to retrieve
configuration settings from the `opensrf.xml` file.

An OpenSRF service spawns one or more children to actually do the work
requested by callers of the service. For every child process an OpenSRF service
spawns, the child process clones the parent environment and then each child
process runs the `child_init()` process (if any) defined in the OpenSRF service
to initialize any child-specific settings.

When the OpenSRF service kills a child process, it invokes the `child_exit()`
procedure (if any) to clean up any resources associated with the child process.
Similarly, when the OpenSRF service is stopped, it calls the `DESTROY()`
procedure to clean up any remaining resources.

=== Retrieving configuration settings ===
The settings for OpenSRF services are maintained in the `opensrf.xml` XML 
configuration file. The structure of the XML document consists of a root
element `<opensrf>` containing two child elements:

  * `<default>` contains an `<apps>` element describing all
OpenSRF services running on this system -- see <<serviceRegistration>> --, as
well as any other arbitrary XML descriptions required for global configuration
purposes. For example, Evergreen uses this section for email notification and
inter-library patron privacy settings.
  * `<hosts>` contains one element per host that participates in
this OpenSRF system. Each host element must include an `<activeapps>` element
that lists all of the services to start on this host when the system starts
up. Each host element can optionally override any of the default settings.

OpenSRF includes a service named `opensrf.settings` to provide distributed
cached access to the configuration settings with a simple API:

  * `opensrf.settings.default_config.get`: accepts zero arguments and returns
the complete set of default settings as a JSON document
  * `opensrf.settings.host_config.get`: accepts one argument (hostname) and
returns the complete set of settings, as customized for that hostname, as a
JSON document
  * `opensrf.settings.xpath.get`: accepts one argument (an
http://www.w3.org/TR/xpath/[XPath] expression) and returns the portion of
the configuration file that matches the expression as a JSON document

For example, to determine whether an Evergreen system uses the opt-in
support for sharing patron information between libraries, you could either
invoke the `opensrf.settings.default_config.get` method and parse the
JSON document to determine the value, or invoke the `opensrf.settings.xpath.get`
method with the XPath `/opensrf/default/share/user/opt_in` argument to
retrieve the value directly.

In practice, OpenSRF includes convenience libraries in all of its client
language bindings to simplify access to configuration values. C offers
osrfConfig.c, Perl offers `OpenSRF::Utils::SettingsClient`, and Python offers
`osrf.set`. These libraries locally cache the configuration file to avoid
network roundtrips for every request and enable the developer to request
specific values without having to manually construct XPath expressions.

== Getting under the covers with OpenSRF ==
Now that you have seen that it truly is easy to create an OpenSRF service, we
can take a look at what is going on under the covers to make all of this work
for you.

=== Get on the messaging bus - safely ===
One of the core innovations of OpenSRF was to use the Extensible Messaging and
Presence Protocol (XMPP, more colloquially known as Jabber) as the messaging
bus that ties OpenSRF services together across servers. XMPP is an "XML
protocol for near-real-time messaging, presence, and request-response services"
(http://www.ietf.org/rfc/rfc3920.txt) that OpenSRF relies on to handle most of
the complexity of networked communications. OpenSRF achieves a measure of
security for its services through the use of public and private XMPP domains;
all OpenSRF services automatically register themselves with the private XMPP
domain, but only those services that register themselves with the public XMPP
domain can be invoked from public OpenSRF clients.

In a minimal OpenSRF deployment, two XMPP users named "router" connect to the
XMPP server, with one connected to the private XMPP domain and one connected to
the public XMPP domain. Similarly, two XMPP users named "opensrf" connect to
the XMPP server via the private and public XMPP domains. When an OpenSRF
service is started, it uses the "opensrf" XMPP user to advertise its
availability with the corresponding router on that XMPP domain; the XMPP server
automatically assigns a Jabber ID (JID) based on the client hostname to each
service's listener process and each connected drone process waiting to carry
out requests. When an OpenSRF router receives a request to invoke a method on a
given service, it connects the requester to the next available listener in the
list of registered listeners for  that service.

The opensrf and router user names, passwords, and domain names, along with the
list of services that should be public, are contained in the `opensrf_core.xml`
configuration file.

=== Message body format ===
OpenSRF was an early adopter of JavaScript Object Notation (JSON). While XMPP
is an XML protocol, the Evergreen developers recognized that the compactness of
the JSON format offered a significant reduction in bandwidth for the volume of
messages that would be generated in an application of that size. In addition,
the ability of languages such as JavaScript, Perl, and Python to generate
native objects with minimal parsing offered an attractive advantage over
invoking an XML parser for every message. Instead, the body of the XMPP message
is a simple JSON structure. For a simple request, like the following example
that simply reverses a string, it looks like a significant overhead: but we get
the advantages of locale support and tracing the request from the requester
through the listener and responder (drone).

.A request for opensrf.simple-text.reverse("foobar"):
[source,xml]
--------------------------------------------------------------------------------
<message from='router@private.localhost/opensrf.simple-text'
  to='opensrf@private.localhost/opensrf.simple-text_listener_at_localhost_6275'
  router_from='opensrf@private.localhost/_karmic_126678.3719_6288'
  router_to='' router_class='' router_command='' osrf_xid=''
>
  <thread>1266781414.366573.12667814146288</thread>
  <body>
[
  {"__c":"osrfMessage","__p":
    {"threadTrace":"1","locale":"en-US","type":"REQUEST","payload":
      {"__c":"osrfMethod","__p":
        {"method":"opensrf.simple-text.reverse","params":["foobar"]}
      }
    }
  }
]
  </body>
</message>
--------------------------------------------------------------------------------

.A response from opensrf.simple-text.reverse("foobar")
[source,xml]
--------------------------------------------------------------------------------
<message from='opensrf@private.localhost/opensrf.simple-text_drone_at_localhost_6285'
  to='opensrf@private.localhost/_karmic_126678.3719_6288'
  router_command='' router_class='' osrf_xid=''
>
  <thread>1266781414.366573.12667814146288</thread>
  <body>
[
  {"__c":"osrfMessage","__p":
    {"threadTrace":"1","payload":
      {"__c":"osrfResult","__p":
        {"status":"OK","content":"raboof","statusCode":200}
      } ,"type":"RESULT","locale":"en-US"}
  },
  {"__c":"osrfMessage","__p":
    {"threadTrace":"1","payload":
      {"__c":"osrfConnectStatus","__p":
        {"status":"Request Complete","statusCode":205}
      },"type":"STATUS","locale":"en-US"}
  }
]
  </body>
</message>
--------------------------------------------------------------------------------

The content of the `<body>` element of the OpenSRF request and result should
look familiar; they match the structure of the <<OpenSRFOverHTTP,OpenSRF over
HTTP examples>> that we previously dissected.

=== Registering OpenSRF methods in depth ===
Let's explore the call to `__PACKAGE__->register_method()`; most of the elements
of the hash are optional, and for the sake of brevity we omitted them in the
previous example. As we have seen in the results of the <<opensrfIntrospection,introspection call>>, a
verbose registration method call is recommended to better enable the internal
documentation. So, for the sake of completeness here, is the set of elements
that you should pass to `__PACKAGE__->register_method()`:

  * `method`: the name of the procedure in this module that is being registered as an OpenSRF method
  * `api_name`: the invocable name of the OpenSRF method; by convention, the OpenSRF service name is used as the prefix
  * `api_level`: (optional) can be used for versioning the methods to allow the use of a deprecated API, but in practical use is always 1
  * `argc`: (optional) the minimal number of arguments that the method expects
  * `stream`: (optional) if this argument is set to any value, then the method supports returning multiple values from a single call to subsequent requests, and OpenSRF automatically creates a corresponding method with ".atomic" appended to its name that returns the complete set of results in a single request; streaming methods are useful if you are returning hundreds of records and want to act on the results as they return
  * `signature`: (optional) a hash describing the method's purpose, arguments, and return value
    ** `desc`: a description of the method's purpose
    ** `params`: an array of hashes, each of which describes one of the method arguments
      *** `name`: the name of the argument
      *** `desc`: a description of the argument's purpose
      *** `type`: the data type of the argument: for example, string, integer, boolean, number, array, or hash
    ** `return`: a hash describing the return value of the method
      *** `desc`: a description of the return value
      *** `type`: the data type of the return value: for example, string, integer, boolean, number, array, or hash

== Evergreen-specific OpenSRF services ==

Evergreen is currently the primary showcase for the use of OpenSRF as an
application architecture. Evergreen 2.6.0 includes the following
set of OpenSRF services:

  * `open-ils.acq` Supports tasks for managing the acquisitions process
  * `open-ils.actor`: Supports common tasks for working with user accounts
     and libraries.
  * `open-ils.auth`: Supports authentication of Evergreen users.
  * `open-ils.auth_proxy`: Support using external services such as LDAP
  directories to authenticate Evergreen users
  * `open-ils.cat`: Supports common cataloging tasks, such as creating,
     modifying, and merging bibliographic and authority records.
  * `open-ils.circ`: Supports circulation tasks such as checking out items and
    calculating due dates.
  * `open-ils.collections`: Supports tasks to assist collections services for
    contacting users with outstanding fines above a certain threshold.
  * `open-ils.cstore`: Supports unrestricted access to Evergreen fieldmapper
    objects. This is a private service.
  * `open-ils.fielder`
  * `open-ils.justintime`: Support tasks for determining if an action/trigger
  event is still valid
  * `open-ils.pcrud`: Supports access to Evergreen fieldmapper objects,
  restricted by staff user permissions. This is a private service.
    objects.
  * `open-ils.permacrud`: Supports access to Evergreen fieldmapper objects,
  restricted by staff user permissions. This is a private service.
  * `open-ils.reporter`: Supports the creation and scheduling of reports.
  * `open-ils.reporter-store`: Supports access to Evergreen fieldmapper objects
    for the reporting service. This is a private service.
  * `open-ils.resolver` Support tasks for integrating with an OpenURL resolver.
  * `open-ils.search`: Supports searching across bibliographic records,
    authority records, serial records, Z39.50 sources, and ZIP codes.
  * `open-ils.serial`: Support tasks for serials management
  * `open-ils.storage`: A deprecated method of providing access to Evergreen
    fieldmapper objects. Implemented in Perl, this service has largely been
    replaced by the much faster C-based `open-ils.cstore` service.
  * `open-ils.supercat`: Supports transforms of MARC records into other formats,
    such as MODS, as well as providing Atom and RSS feeds and SRU access.
  * `open-ils.trigger`: Supports event-based triggers for actions such as
  overdue and holds available notification emails.
  * `open-ils.url_verify`: Support tasks for validating URLs
  * `open-ils.vandelay`: Supports the import and export of batches of
    bibliographic and authority records.
  * `opensrf.settings`: Supports communicating opensrf.xml settings to other services.

Of some interest is that the `open-ils.reporter-store` and `open-ils.cstore`
services have identical implementations. Surfacing them as separate services
enables a deployer of Evergreen to ensure that the reporting service does not
interfere with the performance-critical `open-ils.cstore` service. One can also
direct the reporting service to a read-only database replica to, again, avoid
interference with `open-ils.cstore` which must write to the master database.

There are only a few significant services that are not built on OpenSRF, such
as the SIP and Z39.50 servers. These services implement
different protocols and build on existing daemon architectures (Simple2ZOOM
for Z39.50), but still rely on the other OpenSRF services to provide access
to the Evergreen data. The non-OpenSRF services are reasonably self-contained
and can be deployed on different servers to deliver the same sort of deployment
flexibility as OpenSRF services, but have the disadvantage of not being
integrated into the same configuration and control infrastructure as the
OpenSRF services.

== Evergreen after one year: reflections on OpenSRF ==

http://projectconifer.ca[Project Conifer] has been live on Evergreen for just
over a year now, and as one of the primary technologists I have had to work
closely with the OpenSRF infrastructure during that time. As such, I am in
a position to identify some of the strengths and weaknesses of OpenSRF based
on our experiences.

=== Strengths of OpenSRF ===

As a service infrastructure, OpenSRF has been remarkably reliable. We initially
deployed Evergreen on an unreleased version of both OpenSRF and Evergreen due
to our requirements for some functionality that had not been delivered in a
stable release at that point in time, and despite this risky move we suffered
very little unplanned downtime in the opening months. On July 27, 2009 we
moved to a newer (but still unreleased) version of the OpenSRF and Evergreen
code, and began formally tracking our downtime. Since then, we have achieved
more than 99.9% availability - including scheduled downtime for maintenance.
This compares quite favourably to the maximum of 75% availability that we were
capable of achieving on our previous library system due to the nightly downtime
that was required for our backup process. The OpenSRF "maximum request"
configuration parameter for each service that kills off drone processes after
they have served a given number of requests provides a nice failsafe for
processes that might otherwise suffer from a memory leak or hung process. It
also helps that when we need to apply an update to a Perl service that is
running on multiple servers, we can apply the updated code, then restart the
service on one server at a time to avoid any downtime.

As promised by the OpenSRF infrastructure, we have also been able to tune our
cluster of servers to provide better performance. For example, we were able to
change the number of maximum concurrent processes for our database services
when we noticed that we seeing a performance bottleneck with database access.
Making a configuration change go live simply requires you to restart the
`opensrf.setting` service to pick up the configuration change, then restart the
affected service on each of your servers. We were also able to turn off some of
the less-used OpenSRF services, such as `open-ils.collections`, on one of our
servers to devote more resources on that server to the more frequently used
services and other performance-critical processes such as Apache.

The support for logging and caching that is built into OpenSRF has been
particularly helpful with the development of a custom service for SFX holdings
integration into our catalog. Once I understood how OpenSRF works, most of
the effort required to build that SFX integration service was spent on figuring
out how to properly invoke the SFX API to display human-readable holdings.
Adding a new OpenSRF service and registering several new methods for the
service was relatively easy. The support for directing log messages to syslog
in OpenSRF has also been a boon for both development and debugging when
problems arise in a cluster of five servers; we direct all of our log messages
to a single server where we can inspect the complete set of messages for the
entire cluster in context, rather than trying to piece them together across
servers.

=== Weaknesses ===

The primary weakness of OpenSRF is the lack of either formal or informal
documentation for OpenSRF. There are many frequently asked questions on the
Evergreen mailing lists and IRC channel that indicate that some of the people
running Evergreen or trying to run Evergreen have not been able to find
documentation to help them understand, even at a high level, how the OpenSRF
Router and services work with XMPP and the Apache Web server to provide a
working Evergreen system. Also, over the past few years several developers
have indicated an interest in developing Ruby and PHP bindings for OpenSRF, but
the efforts so far have resulted in no working code. Without a formal
specification, clearly annotated examples, and unit tests for the major OpenSRF
communication use cases that could be ported to the new language as a base set
of expectations for a working binding, the hurdles for a developer new to
OpenSRF are significant. As a result, Evergreen integration efforts with
popular frameworks like Drupal, Blacklight, and VuFind result in the best
practical option for a developer with limited time -- database-level
integration -- which has the unfortunate side effect of being much more likely
to break after an upgrade.

In conjunction with the lack of documentation that makes it hard to get started
with the framework, a disincentive for new developers to contribute to OpenSRF
itself is the lack of integrated unit tests. For a developer to contribute a
significant, non-obvious patch to OpenSRF, they need to manually run through
various (undocumented, again) use cases to try and ensure that the patch
introduced no unanticipated side effects. The same problems hold for Evergreen
itself, although the
http://git.evergreen-ils.org/?p=working/random.git;a=shortlog;h=refs/heads/collab/berick/constrictor[Constrictor] stress-testing
framework offers a way of performing some automated system testing and
performance testing.

These weaknesses could be relatively easily overcome with the effort through
contributions from people with the right skill sets. This article arguably
offers a small set of clear examples at both the networking and application
layer of OpenSRF. A technical writer who understands OpenSRF could contribute a
formal specification to the project.  With a formal specification at their
disposal, a quality assurance expert could create an automated test harness and
a basic set of unit tests that could be incrementally extended to provide more
coverage over time. If one or more continual integration environments are set
up to track the various OpenSRF branches of interest, then the OpenSRF
community would have immediate feedback on build quality. Once a unit testing
framework is in place, more developers might be willing to develop and
contribute patches as they could sanity check their own code without an intense
effort before exposing it to their peers.

== Summary ==
In this article, I attempted to provide both a high-level and detailed overview
of how OpenSRF works, how to build and deploy new OpenSRF services, how to make
requests to OpenSRF method from OpenSRF clients or over HTTP, and why you
should consider it a possible infrastructure for building your next
high-performance system that requires the capability to scale out. In addition,
I surveyed the Evergreen services built on OpenSRF and reflected on the
strengths and weaknesses of the platform based on the experiences of Project
Conifer after a year in production, with some thoughts about areas where the
right application of skills could make a significant difference to the Evergreen
and OpenSRF projects.

== Appendix: Python client ==

Following is a Python client that makes the same OpenSRF calls as the Perl
client:

[source, python] 
--------------------------------------------------------------------------------
include::example$python_client.py[]
--------------------------------------------------------------------------------

NOTE: Python's `dnspython` module refuses to read `/etc/resolv.conf`, so to
access hostnames that are not served up via DNS, such as the extremely common
case of `localhost`, you may need to install a package like `dnsmasq` to act
as a local DNS server for those hostnames.

// vim: set syntax=asciidoc: