- Services
-
Consulting
Consult with us when beginning a new project. We can provide insight on best practices and alternative lower-cost solutions which you may not have considered.
High Performance Computing (Biocluster)
Biocluster is the HPC service for the IGB, a mix of compute nodes and the distributed filesystem mentioned above to help solve essentially every Bioinformatics need. Compute fees are based on the type of node, resources, and amount of time utilized. These fees are billed on a monthly basis are are located in the Biocluster publication or on the website at go.igb.illinois.edu/Biocluster.
Special Projects
The CNRG staff have a diverse set of skills that have allowed for collaboration on programming, workflow, audio/visual and other projects that are outside the normal support supplied to the IGB. If you have a project and would like to investigate if CNRG can assist, please contact us to schedule a meeting.
Computer Support and Analysis
CNRG provides desktop support and configuration suggestions as requested as part of our standard services to the IGB. If you are ordering a new system or are having trouble with your current system, please contact us and we will assist with this purchase.
Desktop Backup
As part of the standard services provided to the IGB, all users have access to 1.5 TB of backup storage and a Crashplan backup license. This will back up your computer constantly throughout the day to backup servers in the IGB and ERML, allowing for data recovery in nearly any conceivable situation. Certain files, like movies and music, are excluded from backup with this service.
Server Backup
All standard servers in the IGB are backed up to tape each day at 5:00PM to our backup service. Every other week, these backup tapes are transferred to an off-site storage facility to provide true disaster recovery.
Poster Printing
CNRG operates a printer capable of printing images of up to 60 inches wide and 100 feet long on a variety of materials, including canvas. The service can be accessed via go.igb.illinois.edu/poster and questions can be directed to posterorders@igb.illinois.edu.
Audio Visual Support and Video Conferencing
CNRG is happy to assist with any AV or video conferencing needs in the IGB. If you have questions or concerns regarding an event that uses AV, please contact us for more details.
Computer Classroom
To facilitate the education of the Genomic and Life Sciences communities of the University of Illinois campus with HPC services, the IGB operates a computer classroom of 48 computer systems with common software installations for use in courses. In addition, a special queue of Biocluster is available for the classroom and unbilled accounts can be created to support the needs of students to learn how to use Biocluster as part of that class.
Web Hosting
The IGB offers two ways for users to host web pages through the IGB. If a user has a private web page they would like to publish they can utilize their home directory on the file server. If a more formal web presence is needed that does not involve significant computation, IGB members can obtain a space for free on the IGB web application servers. For additional information on these processes, please contact us for more details.
HPC Training
Twice a semester CNRG along with HPCBio offers classes on how to use HPC services, with a focus on Biocluster. If you are interested, please ask CNRG for the next date or watch the weekly announcements for the schedule.
Videoconferencing Training
All conference rooms in the IGB have Skype videoconferencing capability and some also allow for videoconferencing through Polycom. If you are interested in being trained on these systems, please contact CNRG.
- Biocluster
-
Cluster Specifications
Storage
1.7 PB of raw storage and growing
Data Storage
Data on Biocluster is not backed up by default and is billed at a rate of $10/TB/month. Should you need small folders to be backed up, this can be arranged for an additional $10/TB/month.
Biodatabase
Many HPC applications require databases to function properly, but databases do not function well on distributed storage systems. To provide a quality database solution for the Biocluster, CNRG created the Biodatabase system. This system is generally available free of charge and is accessible by the worker nodes on the system as well as remote systems.
Large Memory Queue
1 Dell PowerEdge R910 Server 24 Xeon E7540 @ 2.0 GHz cores per node 2 TB RAM per node
- 28 Intel E5-2690 v4 @ 2.6 GHz cores per node
- 768 GB RAM per node
High Memory Queue
2 SuperMicro 1028U SuperServers
Budget Queue
10 Dell PowerEdge R410 Servers 8 Intel Xeon E5530 @ 2.4GHz cores per node 24 GB RAM per node
Default (normal) Queue and Highthroughput Queue
20 Dell PowerEdge R620 Servers 24 Intel E5-2697 @ 2.7GHz CPU cores per node 384 GB RAM per node
- Data Solutions
-
Data Management Plans
In addition to the storage services provided by the IGB, there are also many services offered by campus, and it can become difficult to know which service is the best for your project. To assist faculty in navigating these complexities, CNRG conducts regular audits of Biocluster and is willing to meet with any user to discuss their data flows and the best way to utilize all storage resources on campus. This is a free service to IGB members.
File Server
The File Server is the main enterprise storage platform for the IGB. Providing 131 TB of total storage, this service provides 10 TB of storage to each theme and 50 GB of private storage to each user. This system is backed up daily and the tapes are archived for a minimum of 10 years. The file server is an excellent choice for the primary copy of data.
Biocluster Storage
Biocluster, the primary High Performance Computing (HPC) service at the IGB, has in excess of 800TB of replicated storage available in a highly scaleable distributed storage system. This system is an excellent choice to hold data while it is being analyzed or for ultra-fast transfers of data off-campus. Bills for data utilization (at $10/TB/mo) are generated monthly and delivered through email. This system is not backed up by default and as such is a poor choice for the location of the primary copy of data. Should arrangements need to be made to back up individual folders of Biocluster, that can be arranged for an additional $10/TB/mo.
Box.net
This is a campus service that allows for the storage of FERPA data (grades) and functions similar to Dropbox. Users are given allocations of 50GB, while group folders can be requested that hold up to 2TB. Other key features of Box.net are versioning and the ability to share data securely with users off campus.
Illinois Data Bank
The Data Bank is a campus service that allows for faculty to store up to 2TB of data per year that will eventually be made public on a chosen date. This data is then available for a minimum of five years.
Active Archive
For data that is no longer needed on a regular basis, it may be too expensive to store on Biocluster or too large for the File Server storage limits. This data can be moved to the Active Archive service where for $200/TB/10yrs it can be stored in an easily accessible manner. The Active Archive service utilizes an automatic disk to tape service. This service allows for data to appear as if it is on a regular file system but in actuality is on a tape system, significantly reducing the long-term costs of storage. When data is copied to the Active Archive service, it is actually copied to two tapes, one of which is eventually stored at a secure off-campus location. This makes the Active Archive service an excellent choice for storing the primary copy of data and should be part of every grants Data Management Plan.
NCSA Active Data Storage
The Active Data Storage (ADS) service is meant to serve mid-scale applications for data storage. Offering storage at $96/TB/year for unreplicated storage and $192/TB/year for replicated storage, which is similar to Biocluster, this ADS is an excellent choice for an alternative location of data.
Tape Transfer
In some instances the sheer amount of data generated from a project is too onerous of a cost to store on any service, even the Active Archive service. For these projects, CNRG has a process in place to transfer data directly to tape from the storage system on Biocluster. Although cheaper than the Active Archive service, this system only creates one copy of data on tapes and the tapes are not managed by CNRG. If interested in the tape transfer service, please contact CNRG for further discussion
