CIBR Cluster Co-op

WARNING: The existing CIBR clusters are now ~10 years old. While still functioning properly, apart from a few retired nodes, they are getting old enough that the possibility of a fatal failure is eventually inevitable. Once the clusters die to a sufficient degree that they cannot be cheaply repaired, they will be retired. This may happen without warning, so it is critical that you maintain a local copy of any important data. Note that none of the RAID arrays, including store1, are backed up any more, due to the total failure of store2, so all storage is one major failure away from retirement.

Please note that the CIBR cluster Co-op strives to maximize CPU/cost. On most clusters, staff support is one of the largest expenses. So, while we routinely support issues with hardware, queuing system, operating system problems, etc. we have absolutely no applications support staff. At present, we actually have no BCM support for cluster support staff at all, and Dr. Ludtke's group is handling issues as a community service.

CIBR and a number of specific faculty have sponsored the purchase of several clusters managed for shared use by faculty. CIBR faculty may request moderate allocations for free, as a benefit of CIBR membership (which is free). Faculty who contributed funds directly to the purchase have larger guaranteed allocations.

The clusters currently in operation are:

Name	Cores	Storage	Purchased	Free CIBR Access	Paid Access	Notes
sphere.grid.bcm.edu	960	60 TB + 80 TB	2015	25,000 CPU-hr/qtr	Ludtke, Guan, Waterland
prism.grid.bcm.edu	704	63+88 TB	2013	25,000 CPU-hr/qtr	Wensel, Ludtke, Guan
store	-	350 TB Raid6x2	2013	upon request	-	For inactive data which will still be needed on the clusters. This storage array is available upon request to users of any of the clusters, but is directly accessible only from the clusters to discourage people for using it for general lab backup purposes. It is a RAID6 volume, which gives it good reliability, but every 4-5 years we experience some sort of failure. At one time this storage was backed up in addition to being a RAID6 array, but that is no longer true. If the RAID suffers a critical failure, all data could be lost.

Requesting Access

Free cluster accounts must be requested directly by a PI (tenure track faculty) who must be an established member of CIBR, and such accounts will charge against that PI's allocation. Contact Richard.Sucgang@bcm.edu about CIBR membership. We may ask to discuss your intended use for the cluster before granting accounts, simply to insure that it will not interfere with paid cluster users or violate policy. Please note that cluster security does not meet HIPAA requirements, so no identified patient data may be stored or processed on the clusters.

To request an account, email sludtke@bcm.edu with:

Full name
Requested username
Position (student, postdoc, etc)
Cell Phone Number (for emergencies and initial password)
An email address which must be checked at least daily.

Cluster Usage

Each cluster has independent user accounts, as different faculty contributed different amounts to each. Please click on one of the links above for detailed policies and procedures for each cluster

Individual users are expected to be familiar with the use of Linux clusters in general, and if not, to obtain assistance from their acquaintances, rather than learning by trial and error. Occasionally CIBR offers a workshop to help familiarize people with cluster computing, which will be advertised via normal CIBR mechanisms.

All of the clusters run some version of CentOS Linux. We do not provide any commercial software, but most standard open-source tools are installed, and users are welcome to install commercial or free software within their own accounts. BCM's site license for Matlab is usable on the cluster, but must be installed/licensed in the user's account, and there are some setup issues. Contact Larry Mayran about this for details. We may also be able to install other open-source software system-wide, if you have such a need, please ask.

Details on some of the open-source software we have made available for users is on the Software page.

There is a CIBR Cluster Google Group/Mailing List (https://groups.google.com/forum/#!forum/cibrcluster) which is used for announcements of outages, problems, policy changes, etc. We strongly recommend all CIBR Cluster users join.

In manuscripts, grant proposals, presentations, and other publications

Please acknowledge the assistance from the CIBR Center by including a statement similar to the following: "We gratefully acknowledge the assistance and computing resources provided by the CIBR Center for Computational and Integrative Biomedical Research of Baylor College of Medicine in the completion of this work."

CIBR hosted a Mini-Workshop on cluster computing in 2018: The slides can be downloaded here: workshop2018.pdf Archived video of the presentation is here: Cluster Computing