New system go-live
During 14 months of server specification and purchasing, installation, configuration, testing, training, back and forth communication, and other big or miner preparation works, our most recent version Unicorn has finely finished the operation system transition from Unix to Linux, and the database migration from flat text to Oracle system. Currently, we have 1 production server for our daily staff and patron use and 1 test server used for testing new modules and upgrades. This is the first stage of our launching towards the new system migration, the completed system will consist of 5 Linux and 5 Windows servers, containing four 146.8 GB hard drives and four 72.8 GB hard drives respectively.
The advantages of the new system and database are tremendous and obvious. Adding the characteristics of stable, reliable and powerful, the nature of the open source makes Linux being cost-effective, easily upgradeable, and probable more efficient technical support (Linux has the largest cooperate team and the friendly network in the world). Its most important feature, scalability, will remarkably increase the system workload and decrease the network traffic. At same time, the Oracle relational database provides superior transaction processing, reliable and secure content management, and protection from server failure, site failure, and storage failure. Using the power of Oracle (SQL), all Unicorn data is available for query, including variable length MARC data. The one of the most exciting advantage of Oracle’s database engine is that it offers full support for Unicode, which means we will have our multi-language bibliographic record and our patron can search these items using their original language in the near future! Actually, during the testing period and the first day of new server “go live”, some staff have already given very positive feedback. The staff of bibliographical process department reported that they felt the process speed is much faster comparing the previous system. We believe that more and more benefits will be gradually showing up through the continuing development of our system, such as the full realization of system scalability, the implementing of multilingual bibliographic process and its utilization in the OPAC system.
Looking back at this period of work, we have this conclusion: without the support of every level of our library, the help of all of our team members, the cooperation and collaboration with the vendor and OIT, we could not have achieved this transition and migration in such a smooth way and less painful procedure (certainly, we still had to experience some headache time and struggle for various unpleasant situations).
First of all, we really appreciate the support coming from our management team. The Dean has fully supported us during this process, he contributed his valuable time to give us powerful advice and suggestions. Anne and Doris were extremely helpful in financial services support, always willing to make us feel that our project was a priority. They always completed requested tasks very efficiently and quickly, considering their own heavy workloads.
Secondly, we would like to express our sincere compliment for the library whole team members. Karen Wilson and Lance Gillette of BAM worked very thoroughly and proactively to ensure that our necessary work related tasks would be easily accomplished on the new server. Jim Hassel of Circulation helped us tremendously and aided us to obtain a more broad view of the system by using his years of library experience. Kari Marks has consistently reported us any error and abnormal message that she found by her sharp and alert eyes. Not only that, but her patience and easy going manner made the stressful days easier for us. There are many others who we have not mentioned who also gave various direct and indirect support and assistance. They vigorously helped us test, enthusiastically participated in the training, constructively gave us the feedback, and many others. We apologize for not naming everyone who assisted us in this short report, but we are grateful very much for their contribution.
At the last, we need to say thanks to our vendor and OIT partners for their cooperative approach and patiently and unwearied work. Jared Brumlow and Damon Hunter, SirsiDynix project manager and data conversion specialist, were a pleasure to work with. They communicated effectively during the entire migration project, from scheduling to the physical migration and last minute post-migration modifications and meetings. Maureen English (OIT Oracle Administrator), Anthony Shaw (OIT Windows System Administrator), and Dustin Rice (OIT Linux Systems Administrator) were extremely helpful and cooperative. All three took great efforts to understand the integrated library system and ensure that all pieces of system administration, Oracle installation and backups were completed. Both SirsiDynix and OIT were more than willing to cooperatively work together and listen to our requirements and user needs. Several times during this process SirsiDynix, OIT, and Library IT held conference calls to ensure a smooth transition, we greatly appreciate the efforts that both parties made.
Having such tremendous support and help, we also made a great effort to accomplish this migration tasks. During the process of the migration, there was a lot of communication between library, vendor and OIT department. We consistently made effort to liaise in the follow-through and persistent way, to transfer the huge information promptly and immediately, and make sure the channel of the communication was without any obstacles. Dealing with the transaction from the old system to new one will encounter a lot of new technology and new issues. Our careful preparation and study enabled us possess the good understanding about the related concepts and knowledge, thus enabling us to interpret those high tech messages clearly and accurately between our Vendor and OIT. For supplying the learning-friendly training, we organized, edited and rewrote and training material and made it clear, succinct and easy to understand. Considering the remote libraries real situation and preventing unexpected accident, we gave them alternative choice and more detail advice for the offline module. We notified our staff on time for each change to prevent any unnecessary panic. We tested every step with circumspect attitude. We overcame various inconvenience and difficulties, such as different time zone with the vendor, any misunderstandings between the vendor and OIT, and some abrupt and unannounced requirements. We utilized fully our multitasking capability and jumped our focus from task to task since we still need to deal with vast routine tasks during the migration period. Among these heavy duty and works, what needs to acknowledge is Lisa Smith played the main role and did an excellent job. Her great people skill and flexible manner make our work much easer. Her to be ready at any time attitude enabled our work carrying on smoothly. Her rich experience with integrated library systems is a real asset for our library.
An automation migration is time-consuming, disruptive, and also can be frustrating. During this project, not only we accumulated many experiences in ILS updating, but also increase our skills and techniques in organization, coordinating, and managing library automation project. We also realized that with large projects as this, the only surety is that things will never go entirely as planned. We would like to summarize the lessons that we learnt as the following:
1) Plan and document in detail as much as possible with all involved parties. Since this is a big project and lasted for long term, many people in various departments had been engaging in it. Through the whole project procedure, documentation played a very important role for the tasks that were to be completed and that were transferred to other staff. To be integrated smoothly among staff and departments, full approved document from our early planning stages would have benefited us all greatly. For example, if we had document between the 3 involved parties attending the early stage planning in the beginning, we would not have had some of the communication missing problem that we encountered and cost us some missing tasks in the later migration stage.
2) Ask vendor give the through work flow include each step. SirsiDynix seemed too casual with this project. The migration schedule was very rough and should have included more precise information and planning. If we had acknowledged every step of this migration we would have felt more confident and comfortable during the project. So in the future, we will ask the vendor give us work plan and schedule in detail, and maker sure every step can be fully controlled and traced.
3) Give much more time to OIT and SirsiDynix than it is normal needed. In the future, we should allow outside parties buffer time around their estimated completion dates to allow them time for unexpected emergencies or staff outages that they may face. For example, we originally planned two weeks to test the migration, but we only had 3-4 days because the data migration took longer than SirsiDynix estimated.
4) Test, test, test, deep and more tests. Draft a more detailed test plan and checklist in the future. Consider other aspects that were left out in this plan, such as testing access to the new server from non UA libraries.
5) Preparing alternative, multiple solution/choice. Think ahead and have solutions or work-arounds prepared for unexpected situations or incidents. For example, on the last day of our testing phase, before the production migration, we were notified that Oracle was not being backed up. We received confirmation that this would be resolved before we were prepared to go Live on the new server. We should have prepared a backup plan if OIT was unable to provide an Oracle backup by the specified time.
6) Know more about OIT and vendor, e.g. what can be customized, what is included in contract, and department organization. In the future, we will work with OIT and SirsiDynix to confirm what can be customized, yet still be following our contract rules and regulations. If we had better understood the OIT Organization, we would have been better prepared and the project would have been completed with fewer complications.
What we have done until now is the first step of the long march. A new exploration and exciting moments are waiting for us. We are hoping after finishing the whole system migration, the completed new system will eventually enable us to create a clustered environment allowing us higher availability/failover in case a system component fails. Our configuration plan is to have a production, prep, and test environment, enabling us to completely test an upgrade, patch, or new module on an identical to production, prep server before bringing it up on a production environment. The test server will allow library staff to test new features, modules, reports, etc… at their leisure, without concern of taking up space, time, or corruption of data on a production server. Let us work together to realize it. Thanks!
The Rome’s API
From the Sirsi’s side, we’ve heard that the Rome’API is almost as same as Unicorn’s. It sounds strange. What’s kind expectation that we should have?
SirsiDynix Client Care Email Procedures
ADDRESS
Sirsi Client Care Center <clientcare@sirsi.com>
SUBJECT
Create a New Incident: NEW code
Update Existing Incident: UPDATE incident id
Instructions for Requesting E-mail Support from Client Care http://www.sirsi.com/uhtbin/custinfo/Cccinfo/
1. To request a new incident be created via the Client Care E-mail Support System:
a. Send an e-mail to clientcare@sirsidynix.com
b. Place the keyword syntax of new in the subject line followed by your 5-digit Client Care Code:
Subject: new 54321
c. SirsiDynix has provided an incident form that can be used as a guideline when creating/opening a new incident. Within the body of the
e-mail message please provide a description of your Client Care E-mail Support request.
Description: The contents of the e-mail message are inserted into a new incident that is created for site with a Client Care Code of 54321.
Acknowledgement: A receipt e-mail is sent to either the person who made the request or the group alias if defined. This acknowledgment includes the new Incident ID that has been assigned to your site, and the original description.
2. To request a status on open incidents for your site via the Client Care E-mail Support System:
a. Send an e-mail to clientcare@sirsidynix.com
b. Place the keyword syntax sitestatus in the subject line followed by your 5-digit Client Care Code:
Subject: sitestatus 54321
Description: This request returns a list and brief description of all incidents that have a status of open for site with a Client Care Code of 54321.
Acknowledgment: A receipt e-mail is sent to either the person who made the request or the group alias if defined. This acknowledgment lists incidents that are considered open by the CRM System. The list is composed of the following columns: date the incident was opened, the Incident ID, the incidents status, and the short description. An individual summary of each incident appearing on the list follows. The brief summary includes additional information such as: priority, assigned consultant, and the most recent update associated with the incident.
3. To request a brief status of a specific incident via the Client Care E-mail Support System:
a. Send an e-mail to clientcare@sirsidynix.com
b. Place the keyword syntax of briefstatus in the subject line followed by a specific Incident ID:
Subject: briefstatus 12345
Description: This request returns the original description of the incident, and the most recent update to the incident. With this command you can find out the current status of any incident (either opened, or closed) in the CRM System.
Acknowledgment: A receipt e-mail is sent to either the person who made the request or the group alias if defined. This acknowledgment includes the Incident ID, date opened, priority, assigned consultant, short description, original problem description, and the most recent update for incident 12345.
4. To request a status of a specific incident via the Client Care E-mail Support System:
a. Send an e-mail to clientcare@sirsidynix.com
b. Place the keyword syntax of status in the subject line followed by a specific Incident ID:
Subject: status 12345
Description: This request returns the status of a specific incident (either opened, or closed) maintained in the CRM System.
Acknowledgment: A receipt e-mail is sent to either the person who made the request or the group alias if defined. This acknowledgment includes the date the incident was opened, the date the incident was closed (if applicable), and related updates presented in descending order (most recent to original description).
5. To request an incident be updated via the Client Care E-mail Support System:
a. Send an e-mail to clientcare@sirsidynix.com
b. Place the keyword syntax of update in the subject line followed by a specific Incident ID:
Subject: update 12345
c. Within the body of the e-mail message provide any additional information that you would like to have appended to this incident.
Description: The contents of the e-mail request are added to the incident. When this is done, the E-mail Support System automatically notifies the Client Care representative who is assigned to the incident.
Acknowledgement: A receipt e-mail is sent to either the person who made the request or the group alias if defined. This acknowledgement will include the Incident ID, date opened, priority, assigned consultant, original problem description, and the most recent update for incident 12345.
6. To request that a specific incident be closed via the Client Care E-mail Support System:
a. Send an e-mail to clientcare@sirsidynix.com
b. Place the keyword syntax of close in the subject line followed by a specific Incident ID:
Subject: close 12345
c. At your discretion, within the body of the e-mail message you may provide a reason for closing the incident. This text will be put into the incident resolution.
Description: This request will close the specified Incident ID, when you determine that the incident has been resolved to your approval or that the issue is no longer deemed a problem.
Acknowledgement: The CRM (Customer Relationship Management) System will notify either the person who made the request or the group alias if defined that the incident has been closed. This acknowledgement also contains the Incident ID. The Client Care E-mail Support System will also send a receipt e-mail to either the person who made the request or the group alias if defined. This receipt e-mail contains the Incident ID, priority, assigned consultant, original description and the resolution for the closed incident.
Unicorn In Action 