8 High Availability

High availability (HA) can be defined as follows. The system is designed for highest requirements in terms of performance and reliability. Several platform capabilities allow easy scaling without downtimes.

From version 2.2 Intershop Order Management is ready for High availability.

The following sections describe a tested and working approach to enable IOM to be HA.

8.1 Symmetric HA Nodes

High availability can be provided by using symmetric high availability nodes. Therefore every server node consists of:

• A Wildfly application server acting as frontend server
• A Wildfly application server acting as backend server
• An FTP-Server with access to synchronized file-system

The configuration of all high availability nodes is identical, so every server node is able to replace any other. As a result, the entire system can answer requests as long as at least one node exists.

The following diagram shows an example of high availability nodes. GlusterFS can be used to replicate data between FTP servers and backend servers. IOM Watchdog for managing the availability of application servers is not mentioned here.

8.2 Load Balancer

If two or more frontend servers are used, a load balancer has to be placed in front of them. This load balancer has to support:

• Session stickiness
• Session failover for undertow-subsystem
• Health check requests

8.2.1 Session Stickiness

In general, all session data are persisted within the database. But there are some transient data stored at the session, which are used in AJAX requests only. The called page prepares some information stored at the session and AJAX calls embedded into the response of the page are accessing them. This mechanism only works if page- and AJAX-requests are both handled by the same frontend application server.

Wildfly Application Server supports sticky sessions by appending a route information to the session cookie. The route is simply a unique identifier of the server that should handle the session. As long as this server is available, the load balancer has to send all requests of this session to the same server. To guarantee this behavior, the load balancer has to support this kind of session stickiness.

Both, frontend application server and load balancer have to use identical routes. At the application server side the route is configured by the following property, located at $OMS_ETC/system.std.frontend.properties:

# Make sure the "route" property at the load balancer uses the same ID as set here for "instance-id".
# The value assigned to instance-id will be appended to the session-ID. With the help of this information
# the load balancer (e.g., mod_prox_balancer) realizes session stickiness.
/subsystem=undertow: instance-id="${installation.SERVER_ID}"

This example shows the default configuration of IOM. instance-id is set to the value ${installation.SERVER_ID}, which itself is defined as SERVER_ID in $OMS_ETC/installation.properties.

8.2.2 Session Failover

If one of the frontend servers disappears, e.g., for maintenance, due to a hardware failure etc., the load balancer has to be able to send all incoming requests dedicated for this frontend server to the remaining frontend servers.

8.2.3 Health Check Requests

The load balancer should support health check requests. The health check requests support the load balancer in deciding which frontend application server to use and which not, e.g., if the frontend server is currently redeployed, is unable to connect to the database or if it is unable to connect the FTP server, then the load balancer should not use the according frontend server any longer. 

8.2.4 Load Balancer - Using Apache HTTP Server Configuration

You can use any load balancer you want, as long as it is able to fulfill the requirements described in the sections above.

If you choose Apache HTTP server to be used as load balancer, the combination of mod_proxy_balancer and mod_proxy_hcheck is able to fulfill the requirements. The example below shows a configuration snippet of mod_proxy_balancer and mod_proxy_hcheck made for two frontend application servers. The according machines have IPs 10.10.10.1 and 10.10.10.2 with OMT application listening on port 8180. The value of route uses the default setting of SERVER_ID, which is set in $OMS_ETC/installation.properties to $(hostname)_$(OMS_SERVER_TYPE). The health check URL is set to /monitoring/services/health/status for both servers. The interval for checking the health status is set to 10 seconds.

# load modules required for reverse proxy
LoadModule proxy_module modules/mod_proxy.so
LoadModule proxy_http_module modules/mod_proxy_http.so
LoadModule proxy_balancer_module modules/mod_proxy_balancer.so
LoadModule proxy_hcheck_module modules/mod_proxy_hcheck.so
LoadModule watchdog_module modules/mod_watchdog.so
LoadModule lbmethod_byrequests_module modules/mod_lbmethod_byrequests.so
LoadModule lbmethod_bytraffic_module modules/mod_lbmethod_bytraffic.so
LoadModule lbmethod_bybusyness_module modules/mod_lbmethod_bybusyness.so
LoadModule slotmem_shm_module modules/mod_slotmem_shm.so

# make sure, not to act as forward proxy
ProxyRequests Off

# do not rewrite Host-Header when forwarding request
ProxyPreserveHost on

# define proxy with balancer-members and settings
# make sure, route is identical to the settings made at the frontend-application-servers
<Proxy balancer://omscluster>
    BalancerMember http://10.10.10.1:8180 max=5 timeout=40 retry=40 acquire=30000 route=fe1.iom.com_frontend hcmethod=GET hcinterval=10 hcuri=/monitoring/services/health/status
    BalancerMember http://10.10.10.2:8180 max=5 timeout=40 retry=40 acquire=30000 route=fe2.iom.com_frontend hcmethod=GET hcinterval=10 hcuri=/monitoring/services/health/status
</Proxy>

# enable balancer-manager, avoid proxying
<Location /balancer-manager>
  SetHandler balancer-manager
  ProxyPass !
</Location>
# enable server-status, avoid proxying
ExtendedStatus on
<Location /server-status>
  SetHandler server-status
  ProxyPass !
</Location>

# proxying everything else
ProxyPass / balancer://omscluster/ stickysession=JSESSIONID
ProxyPassReverse / balancer://omscluster/ stickysession=JSESSIONID

8.3 Frontend Application Server as Part of HA Node

Frontend servers do not hold any state information within the file system and the business processes running in frontend servers can be interrupted at any time. Therefore you can start, stop, create or destroy them whenever you like.

8.3.1 IOM Watchdog

In addition to the health checks made on the frontend servers controlling the behavior of the load balancer, it is recommended to run the frontend server under control of the server monitoring tool, i.e., IOM Watchdog. Additionally, the Watchdog provides the ability to restart the server whenever it becomes unhealthy or the process dies unexpectedly. Please see Guide - IOM Watchdog 2.2 - 2.11 and Health Check Requests and Cluster Status for more information.

Since traffic to frontend servers is already controlled by health check requests made by the load balancer, which can react quite fast (it can enable/disable a frontend server within some seconds), the reaction of the watchdog on unhealthy frontend servers should be more lazy, e.g., react after 5 minutes in unhealthy state. The according configuration is made in $OMS_ETC/watchdog.properties of the frontend server.

# number of seconds before failed health-checks lead to restart of the
# watched process
watchdog.healthcheck.timeout = 300

Frontend servers are working in parallel. The failover feature of watchdog has to be disabled for this type of server. The according configuration is made in $OMS_ETC/watchdog.properties of the frontend server.

# switch on/off failover functionality. Has to be set to true, when watchdog
# controls a backend-server. Has to be set to false, when a frontend-server
# is controlled.
watchdog.failover.enabled = false

8.3.2 Port Offset

Since frontend and backend server combined as HA node would compete for the same ports, the port configuration or the address binding of one of the servers must be changed. The easiest and recommended solution is setting JBOSS_PORT_OFFSET in $OMS_ETC/installation.properties of the frontend server to a value of 100. Setting this property this way adds 100 to all port numbers of the frontend server, e.g., the port of the undertow subsystem (embedded web-server) changes from 8080, the default value, to 8180, Wildfly administration port changes from 9990 (default) to 10090.

When managing the frontend server via jboss-cli.sh, it is important to always pass the correct address and port to the program. IOMs set_env.sh helps to provide according variables JBOSS_BIND_ADDRESS and JBOSS_MGMT_PORT. When using jboss-cli.sh you have to pass them as: --controller=$JBOSS_BIND_ADDRESS:$JBOSS_MGMT_PORT.

8.4 Backend Application Server as Part of HA Node

Backend servers are stateful. There are some measures to deal with this statefulness:

• Identical impex jobs must not run on different backend servers at the same time.
• Directories containing stateful data (e.g., import or export files) are synchronized to all backend servers.

8.4.1 IOM Watchdog

In contrast to the frontend servers, the usage of the IOM Watchdog for backend servers is not optional. The most important purposes of the watchdog for backend servers are:

• Make sure that an unhealthy server does not process job tasks any longer.
• Make sure that only one backend server is active.

The Watchdog starts the backend server only if no other backend server is already active. IOM Watchdog stops the backend server process in case it becomes unhealthy or loses its active status. Hence, Watchdog for backend servers has the same purpose as the health checks of the load balancer on frontend servers. It guarantees that an unhealthy backend server will be taken out of the system. Exactly as the health checks made by the load balancer, an unhealthy backend server has to be deactivated within some seconds. Therefore it is recommended to use a much shorter health check timeout for backend servers. The according configuration is made in $OMS_ETC/watchdog.properties of the backend server.

# number of seconds before failed health-checks lead to restart of the
# watched process
watchdog.healthcheck.timeout = 20

Watchdog uses a centralized repository (database) where all Watchdogs are negotiating the currently active backend server. All Watchdogs are constantly checking this database. If the formerly active backend server has lost its active state, e.g., because it became unhealthy, the next watchdog will take over. It claims the active state and starts another backend server on another node. The failover feature has to be enabled for backend servers in $OMS_ETC/watchdog.properties.

# switch on/off failover functionality. Has to be set to true, when watchdog
# controls a backend-server. Has to be set to false, when a frontend-server
# is controlled.
watchdog.failover.enabled = true

For details please see Guide - IOM Watchdog 2.2 - 2.11.

8.4.2 Job Scheduling

Cluster jobs only exist on the backend server. They require no special attention in regard to HA with the current approach using only one active backend server. Please see Guide - Intershop Order Management - Job Scheduling | Clustered Jobs for more information.

8.4.3 Data Synchronization

Some directories of the backend server are containing stateful data. These directories have to be synchronized across all backend servers. The following directories of backend servers have to be synchronized:

• $OMS_VAR/communication/messages
• $OMS_VAR/importarticle
• $OMS_VAR/jobs

It is recommended to use a kind of clustered files system for synchronization (e.g., GlusterFS). 

8.5 FTP Server as Part of HA Node

IOM uses two FTP servers for internal communication between frontend and backend servers. The HA node combines both FTP servers to a single server having two accounts. The FTP server as part of an HA node is used by the local frontend and local backend server only. All IOM application servers have to use the following configuration in $OMS_ETC/cluster.properties:

/system-property=is.oms.media.host: "localhost"
/system-property=is.oms.pdf.host: "localhost"

The directories, the FTP server accounts are using, are synchronized between all HA nodes. For synchronization a clustered file system should be used, since all nodes are equal in this case. There is no single node with a master role.

The following directories have to be synchronized between all frontend servers:

• $OMS_VAR/pdfhost
• $OMS_VAR/mediahost

Checking the functionality of the FTP server is part of the health check, see Concept - IOM Server Health Check.

Depending on your infrastructure, you can choose different solutions to provide a highly available FTP service for IOM:

• If you already have a highly available FTP server, you can use it. You can configure all IOM application servers to point on this FTP server.
• If you already have a highly available network file system, you can use multiple FTP servers, all accessing the network file system. Every IOM application server should access a locally installed FTP server which uses the HA network file system to share/synchronize data.

8.6 Health Check Requests and Cluster Status

8.6.1 Health Check Requests

All application servers provide a health check that can be requested using the URL at /monitoring/services/health/status. It responds with HTTP status code 200 if the application server is healthy, otherwise it responds with 5XX.

To ease error analysis, the content delivered by the health check URL contains further information about processed checks. These information is provided in a machine-readable format (JSON), which can also be easily understood by humans.

Health checks require the deployment of oms.monitoring-app-2.x.x.x.war.

Health checks of FTP servers can be enabled or disabled. The according settings are realized by the two properties:

/system-property=is.oms.media.healthcheck: "enabled"|"disabled"
/system-property=is.oms.pdf.healthcheck: "enabled"|"disabled"

See Concept - IOM Server Health Check for more information.

8.6.2 Cluster Status

For an overview about all application servers within a cluster, each health check request updates the current status of the application server within the database. The cluster status can be seen using the URL /monitoring/services/health/clusterstatus. Basic authentication is required. The user also requires the permission Basic application management assigned at root organization.

See Concept - IOM Server Health Check for more information.

8.7 High Availability Database

Frontend servers and backend servers are all connecting to the PostgreSQL database.

For high availability the database and the connection to it also have to support HA:

• Database reconnect
• PostgreSQL HA cluster

8.7.1 DB Reconnect

To provide HA, the application servers are able to reconnect to the database without the need of a restart.

To work properly, invalid connections must be recognized and evicted from the pool. The xa-datasource configuration defines how this happens. We recommend to use the background validation checker rather than the validate-on-match method to reduce the check overhead. Moreover, the timeout's configuration parameters may influence the reconnect behavior (old connections might not get evicted as long as the timeouts are not reached).
Please see redhat's Administration and Configuration Guide for more information about datasource configuration.

xa- datasource configuration example
# The pool size depends on the number of application servers and the database server ressources
min-pool-size="30"
max-pool-size="80"
pool-prefill="true"

# timeouts
set-tx-query-timeout="true"
query-timeout="3600"
blocking-timeout-wait-millis="3000"
idle-timeout-minutes="60"

#connection validation
validate-on-match="false"
background-validation="true"
background-validation-millis="20000"
exception-sorter-class-name="org.jboss.jca.adapters.jdbc.extensions.postgres.PostgreSQLExceptionSorter"
valid-connection-checker-class-name="org.jboss.jca.adapters.jdbc.extensions.postgres.PostgreSQLValidConnectionChecker"

# ressources that are unused or not supported by IOM
interleaving="false"
pad-xid="false"
wrap-xa-resource="false"
same-rm-override="false"
share-prepared-statements="false"

8.7.2 PostgreSQL HA Cluster

IOM supports access to PostgreSQL HA clusters, but always has to be connected to the master database.

A PostgreSQL HA cluster usually consists of one master server and one or more hot-standby servers. The master server is the only one that is allowed to change data. An additional witness server is needed by the failover process when the total number of servers (master + standbys) is odd.

During the failover, the IOM application must be redirected to the new master. One solution is to add a proxy layer between the IOM application servers and the Postgres HA cluster. This proxy layer can be realized by PgBouncer. PgBouncer has to be reconfigured on the fly (without restart) whenever the current master changes. PgBouncer being a connection pool, can also be used to limit the number of connections to PostgreSQL. More than one instance of PgBoucer should be defined to avoid single points of failures.

The IOM database connection address is defined in $OMS_ETC/cluster.properties by the property is.oms.db.hostlist which supports a number of one or more database host addresses. For more information please see Guide - Setup Intershop Order Management 2.10

For more information about PostgreSQL HA clusters see http://repmgr.org and https://pgbouncer.github.io

8.8 Clock Synchronization

All hosts of the cluster should use a clock synchronization service (daemon).