The EMBL Nucleotide Sequence Database () at the EMBL European Bioinformatics Institute, UK, offers a large and freely accessible collection of nucleotide sequences and accompanying annotation. The database is maintained in collaboration with DDBJ and GenBank. Data are exchanged between the collaborating databases on a daily basis to achieve optimal synchrony. Webin is the preferred tool for individual submissions of nucleotide sequences, including Third Party Annotation, alignments and bulk data. Automated procedures are provided for submissions from large-scale sequencing projects and data from the European Patent Office. In 2006, the volume of data has continued to grow exponentially. Access to the data is provided via SRS, ftp and variety of other methods. Extensive external and internal cross-references enable users to search for related information across other databases and within the database. All available resources can be accessed via the EBI home page at . Changes over the past year include changes to the file format, further development of the EMBLCDS dataset and developments to the XML format.

Absrmcl-Performance modeling is an important topic in capacity planning and overload control for web servers. We present an M/GNK*PS queueing model of a web server. The arrival process of HTTP requests is assumed to he Poissonian and the semce discipline is processor sharing. The total number of requests that can he processed at one t i e is l i t e d to K. We obtain closed form expressions for web sewer performance metrics such as average response time, throughput and hlofking probability. The average of the service time requirement and the limit of the number of requests being Served are model parameters. The parameters are estimated by maximizing the log-likelihood function d the measumd average response time. Compared to other models, our model is conceptually simple and it is easy to estimate model parameters. The model has been validated thmngh measurements in our lab. The performance metrics predicted by the model fit well to the experimental outcome. I. INTRODUCTIONPerformance modeling is an important part of the research area of web servers. Without a correct model of a web server it is difficult to give an accurate prediction of performance metrics. A validated model is the hasis of web server capacity planning, where models are used to predict performance in different settings, see Hu et al. 111 or M e n a d and Almeida 121.

The EMBL Nucleotide Sequence Database () at the EMBL European Bioinformatics Institute, UK, offers a comprehensive set of publicly available nucleotide sequence and annotation, freely accessible to all. Maintained in collaboration with partners DDBJ and GenBank, coverage includes whole genome sequencing project data, directly submitted sequence, sequence recorded in support of patent applications and much more. The database continues to offer submission tools, data retrieval facilities and user support. In 2005, the volume of data offered has continued to grow exponentially. In addition to the newly presented data, the database encompasses a range of new data types generated by novel technologies, offers enhanced presentation and searchability of the data and has greater integration with other data resources offered at the EBI and elsewhere. In stride with these developing data types, the database has continued to develop submission and retrieval tools to maximise the information content of submitted data and to offer the simplest possible submission routes for data producers. New developments, the submission process, data retrieval and access to support are presented in this paper, along with links to sources of further information.