Main tasks:
* Warns the user when the backup is running out of space ([[#Quota |quota]]).* Warns the user if the backup wasn't executed at the planned time ([[#Schedule |schedule]]).* Informs the user if the backup was executed, but finished with errors ([[#Unsuccessful |unsuccessful]]).
* If the Backup Client is too old, inform the user (upgrade) (not implemented yet).
* Tell the user, that they have a backup account, but it was never used (no backup) (not implemented yet).
= Tasks =
== Quota ==
If the used quota in percentage is bigger than the threshold, the appropriate error message is sent (see [[#Who_will_receive_notification_mails| who will be informed]]). The threshold is variable value which is read from the accounts backend (LDAP directory) entry. (The reseller and/or the customer can change the threshold for the given account using the [[stoney_backup:_Self-Service#Modify_backup_account | selfcare web-interface ]] (not implemented yet).) The script [[stoney_backup:_Service_Software#writeAccountSize.pl | writeAccountSize.pl]] (see [[#Source Code]]) is responsible for this task.
== Schedule ==
Each (more or less up-to-date) backup client (developed by [http://www.stepping-stone.ch stepping stone GmbH]) uploads scheduling information to the server when backing up a computer. According to the scheduling information, the server can check if a backup was executed at the correct time or not. If not an appropriate error message is sent (see [[#Who_will_receive_notification_mails | who will be informed]]). (The reseller and/or customer can define after how many not executed backups he or she wants to be informed using the [[stoney_backup:_Self-Service#Modify_backup_account | selfcare web-interface]] (not implemented yet).)
The script [[stoney_backup:_Service_Software#scheduleWarning.pl | scheduleWarning.pl]] (see [[#Source Code]]) is responsible for this task.
== Unsuccessful ==
Each (more or less up-to-date) backup client (developed by [http://www.stepping-stone.ch stepping stone GmbH]) uploads the status of the backup to the server after backing up a computer. According to this status information, the server can check if the last backup was successful. If not an appropriate error message is sent (see [[#Who_will_receive_notification_mails | who will be informed]]). (The reseller and/or customer can define after how many not successful backups he or she wants to be informed using the [[stoney_backup:_Self-Service#Modify_backup_account | selfcare web-interface]] (not implemented yet).)
The script [[stoney_backup:_Service_Software#scheduleWarning.pl | scheduleWarning.pl]] (see [[#Source Code]]) is responsible for this task.
= Who will receive notification mails =
If one (or more) of the [[#Tasks | tasks]] result in sending an error message, the notification scripts collects the mail addresses to whom it must send the message(s) from the LDAP directory backend in the following way:
== Customer / User ==
# Check if the customer / user must be informed
#* If <code>sstBackupWarningOn</code> of the accounts backend (LDAP directory) entry is set to <code>TRUE</code> the customer / user has to be informed
#* If <code>sstBackupWarningOn</code> of the accounts backend (LDAP directory) entry is set to <code>FALSE</code> the customer / user does not have to be informed
# Check if the account itself has some E-Mail addresses attached
#* Check if the LDAP attribute <code>mail</code> exists for the accounts backend (LDAP directory) entry AND <code>sstNotificationWarningMedium</code> is set to <code>mail</code>.
#** If yes, send the mail to all (the attribute <code>mail</code> is multivalued, see [[stoney_backup:_OpenLDAP_directory_data_organisation#Backup_Account_Example | OpenLDAP directory data organisation]]) addresses specified in the attribute <code>mail</code>
#** If not, check to whom the accounts belongs
# Check to which person the backup accounts belongs
#* Read the attribute <code>sstBelongsToPersonUID</code> and search the given person UID in the backend (LDAP directory).
#* Read the attibute <code>mail</code> from the persons bachend (LDAP directory) entry retrieved form the previous serach
#* Send the mail to all (the attribute <code>mail</code> is multivalued, see [[stoney_backup:_OpenLDAP_directory_data_organisation#Backup_Account_Example | OpenLDAP directory data organisation]]) addresses specified in the attribute <code>mail</code>
(The reseller and/or customer can define the values of <code>sstBackupWarningOn</code>, <code>sstNotificationWarningMedium</code> and <code>mail</code> using the [[stoney_backup:_Self-Service#Modify_backup_account | selfcare web-interface]] (not implemented yet).)
== Reseller ==
# Check if the reseller wants to be informed
#* If the given problem (see [[#Tasks]]) is part of the attribute <code>sstNotificationWarning</code> in the resellers notification-settings backend (LDAP directory) entry, the reseller has to be informed.
#* If the given problem (see [[#Tasks]]) is not part of the attribute <code>sstNotificationWarning</code> in the resellers notification-settings backend (LDAP directory) entry, the reseller does not have to be informed.
# Send the notification mail (reseller version) to all (the attribute <code>sstMailTo</code> is multivalued, see [[stoney_backup:_OpenLDAP_directory_data_organisation#Backup_Reseller_Backup_Notification_Settings | OpenLDAP directory data organisation]]) addresses specified in the attribute <code>sstMailTo</code>
== Graphical ==
[[File:Notification-receiver-workflow.png|650px|thumbnail|none|Figure 1: Who will receive notification mails]]
= Template distribution =
As the templates should be editable for resellers, they are stored on the server which is running the selfcare web interface. On a default stoney cloud setup this will be the [[Primary-Master-Node | primary master node]]. However as described on the [[stoney_cloud:_Notification_Architecture#File_System | notification architecture page]], the templates need to be present on server where the notification process takes place. As we do not have shared storage resources for the primary master node and VMs, the solution is to regularly (once a day) copy the templates from the primary master node to where they should go (for example the backup server):
/etc/cron.d/distribute-templates
<pre>
# Distribute selfcare templates
00 04 * * * /root/distributeTemplates.sh | logger -t distributeTemplates
</pre>
/root/distributeTemplates.sh
<source lang="bash">
#!/bin/bash
readonly backup_server="<Server>"
readonly username="<user>"
readonly identity_file=/root/.ssh/id_rsa_demo
readonly themes_source_directory=/var/www/selfcare/htdocs/themes/
readonly themes_destination_directory=/var/www/selfcare/htdocs/themes/
readonly files_from="/root/distribute_backup_templates.txt";
rsync -vrlt --delete -e "ssh -i ${identity_file}" --files-from ${files_from} ${themes_source_directory} ${username}@${backup_server}:${themes_destination_directory}
</source>
/root/distribute_backup_templates.txt
<pre>
<reseller1-themes-direcory>/templates/services/<servce>/
<reseller2-themes-direcory>/templates/services/<servce>/
</pre>
For example:
<pre>
selfcare.reseller.org/templates/services/backup/
selfcare.another-reseller.org/templates/services/backup/
</pre>
= Source Code =
The source code is located in our GitHub Repository: