Confluence Backup Strategy for large instances

Recently I have discovered an tool to speed up backing up of large Confluence instances.

Rsync is similar to ftp (file transfer protocol) but it is much faster because it only transfers the modified portions of the files.

By using Rsync or cwSync (for windows) in place of the default , it addressed the following issues:

• speed - the backup takes minutes rather than hours now
• memory consumption - it does not require huge memory to be allocated to the JVM
• responsiveness - people can still surf on Confluence during the backup process

KM Asia 2007

Yeah! I am going to attend KM Asia 2007 tomorrow. I am so excited with the opportunity to meet and learn from like-minded people tomorrow. Do drop a comment if you are attending as well.

An insight on Email for corporate needs (2)

Keeping track of assigned tasks

Emails are one of the most natural tools to use for keeping track of assigned tasks. Tasks are usually a consequence from planning and discussion and email is one of the channels for discussions. If there is a deadline for the task, reminders can also be set easily.

What are the advantages

• There is no need to key in the tasks into a separate spreadsheet or tasks
• Does not incur extra overheads
• Reminder can be used without separate programs or more reminder emails because the email client is launched everyday
• Updates to the tasks can be conveniently done by just replying to the email
• Email is a flexible and powerful tool with attachments capabilities

What are the disadvantages

• No big picture on the current workload
• Difficult to measure the amount of work being assigned and completed
• Tasks being assigned are piecewise. Difficulty in consolidating the total effort spent by the team or the project
• Difficulty in consolidating the progress of the tasks if the multiple tasks are batched out in an email
• Difficulty in visualizing any dependencies and linkages in the tasks
• Difficulty in setting the priority of the tasks
• Possible to losing track of tasks being assigned if the assignees do not respond
• Tasks are mixed together with other types of emails
• Difficulty in categorizing and tracking all the tasks assigned out
• Possibility that someone will redo the same task again
• The eventual completed task and other relevant useful information may not be available for others to reference when they need to do the similar task again.

An insight on Email for corporate needs

Recently there are a lot of discussions about whether email will be replaced by the Enterprise 2.0 tools like RSS Reader, Wikis, Blogs and online messengers. The business week has a brought up the debate. It is an interesting topic for everyone since emails are used vastly for business and personal purposes and what will happen if we cannot use emails anymore.?

I shall share in a series of blog posts on the different aspects of using emails for my work and list down their pros and cons. These are the few that popped out my mind at the moment.

1. For conducting electronic discussions
2. Keeping track of assigned tasks
3. Getting notified for meeting requests
4. Keep record of important information
5. Getting notified for updates (mailing lists, notifications)
6. Non important communications

If you have other uses for email, do share with me. Finally I will evaluate the other tools against using email to see if they can help us to improve our productivity.

1. For conducting electronic discussions

With the specialization of skills and increased complexity of work, most of us are required to collaborate in a team to get things done. One of the activities of collaboration is discussion. It may take place at the following levels:

• across functional units
• within the same team
• across hierachary levels

While face to face communication is the most ideal mode of communication, electronic communication is being used because of the following reasons:

• geographical distance
• not everyone can meet at the same time
• scope of discussion too small to warrant for face to face discussion
• immediate response not possible

What are the advantages

1. Confidentiality. Ability to limit the discussion group easily. Members can be added or removed
2. Cheap. What happens if emails are being charged?
3. Ease of use and user friendly
4. Allow the point to get through (via textual and graphical attachments and formatting)
5. Allow storage and retrieval subsequently
6. Relative prompt response depends on the users' work culture
7. Allow flexible categorization of emails into folders
8. Allow recall of emails
9. Allow asychronous discussion
10. Large user base. Almost everyone has an email account and some have more than one.

What are the disadvantages

1. Discussions cannot be tracked easily with multiple threads being spinned
2. Other interested personnel cannot join in the discussion easily without getting invited
3. Cannot filter out the important points of discussions
4. Loss of context when emails are not read in the proper sequence and thus spending more time to re-read the entire email for (Gmail attempts to tackle this)
5. Time consuming to read the entire email (Gmail attempts to tackle this)
6. Distraction when emails get popped
7. A need to file the emails in
8. Duplicated storage of the same information
9. Information is being lost when the mailbox corrupts or PC is reformatted.
10. No control over the eventual audiences of the emails (when it get forwarded around)

Organisational Wiki Adoption

Mike Cannon-Brookes gave a presentation on Organisational Wiki Adoption. The slides covered many interesting points and ideas on wiki adoption. It's a pity that I did not have the opportunity to listen to the presentation.

Plug-in policy for Confluence

Someone asked for comments on plug-in policy for enterprise installation in the Confluence mailing list. So it is time for me to share my experiences on Confluence administration.

While the plugin architecture adopted by Confluence has spawned a large number of useful extensions, there are a few potential concerns to be addressed which has been amplified by the recent 2.6.0 release.

• Whether plugins are going to be supported and continued to be updated (esp after api changes)
• whether the plugin is properly designed
• whether the plugin has been tested adequately
• whether the plugin can be extensible
• whether the plugin is safe to use (without viruses/trojans)
• the amount of effort to test the plugin for each upgrade

Although we did not come up with a official policy at the moment, there are a few criteria I am using for selection.

Functionality

• Is the plugin really useful? If it is going to be heavily used, then consider it.
• User friendliness of the macro - If it is difficult to use, then most likely it is not going to be used.
• Can the functionality achieved by existing functionality like reporting plugin or other macros?
• whether any data is being stored with the macro? If the macro only extract data or reformat the data?

Support

• What will happen if the support for the macro is stopped? Any data loss or tedious data patching?
• Is the plugin developed by big company or established developers (i.e Atlassian, Adaptavist, Customware, Gliffy, ....)
• Whether there are more than 1 release? Beta releases are usually dangerous for production environment, and subsequent release indicates likelihood of future support.
• Is the plugin compatible with the current and subsequent versions (i.e http://confluence.atlassian.com/display/CONFEXT/Tagcloud+Macro)
• Is it something very complex? If it is too complex, it will be hard to migrate or extend in future
• Is the source code being provided (to cater worst case scenario to fix personally)

Stability

• Any comments or bugs raised by users in the product page in the Confluence site and the plugin's jira site?
• Whether it is easy to understand how the functionality of the plugin is achieved. It is not recommended to use something that you are not sure

Security

• Whether the plugin requires internet connection? (i.e editgrid)
• Whether the plugin require special permission to run (eg. administrators)
• Whether the plugin will break existing permission access?

Cost

• The type of license. Whether payment is required? What happens if the price of licensing has increased and the data cannot be migrated

MIT Sloan Management Review

With the advancement in IT, it has indeed provided the impetus for the creation of 24 hrs knowledge factories. Imagine the benefits that tasks can be completed faster without having staff to put in additional overtime and any lag time. Tasks can be continued in another time zone where the first team goes home to rest. And the entire team will work at full efficiency with adequate rest and balanced lifestyle.



If you are interested, read on at Expanding the 24-Hour Workplace



More articles can be found at MIT Sloan Management Review