Marco's Blog! > A cell counting demo using BOSCO and OSG Connect

This post describes a demo that I ran few months ago on Open Science Grid using BOSCO and OSG-Connect.
Let's assume I have a lot of scientific data to analyze, e.g. I want to count the cells in many images stored on my computer, and I don't have the computing power to complete in a timely manner. Opportunistic resources on Open Science Grid and some tools available to its community can help me get the task done. In this tutorial I will show:
  • how to register on OSG Connect to get access to OSG resources 
  • how to easily move your data so that it can be available on the Web and used by your jobs 
  • how to install and use BOSCO to submit the python jobs and retrieve the results on your laptop or workstation 

Here is a recording and below is an extended version of the original tutorial. Although OSG is still active, the original tutorial page and some of the resources described below may not be available any more.




The task

I have several pictures like this:
Figure 1.  Sample image
These pictures look a bit like biological samples where cells (clusters of different luminosity) are visible on top of a background. And I'd like to count the "cells" into the picture. This is done by a clusterization process illustrated in Figure 2 and in this tutorial. The image on the left is just adding colors but has the same information of the black and white one. Using the intensity of all pixels the algorithm evaluates a threshold and builds a mask (second image) that allows to separate the important elements from the background. Finally the distinct elements are separated and counted.

Figure 2.  Example of clusterization

Register on OSG Connect

Before using OSG you have to get access to the resources. This page explains how to join OSG. Here are the practical steps:
  1. Go to the OSG Connect web site and register by clicking on "Log In or Register" (upper right corner) and following the instructions. Here a document explaining the registration
  2. Join a project or create a new project (e.g. one describing your research activity Test-CellCounting).
  3. Login on the submit host login.osgconnect.net and start submitting your jobs. Here is a quickstart guide but you'll feel right at home if you are familiar with HTCondor. Just remember to specify a ProjectName in your jobs or to add a $HOME/.osg_default_project file with your default project name. And there is much more: ConnectBook is a collection of guides about OSG Connect.

Transfer the files

In the submit host the directory ~/data/public is exported on the Web by stash, the OSG Connect data server, as https://stash.osgconnect.net/+USER_NAME/ (where USER_NAME is your user name on OSG Connect)and you can use it to move your data. See Access Stash remotely from your job using HTTP for an example.
I will use this feature and transfer my files there using Globus Online.

Install and setup BOSCO

Download & Install Bosco

This is a very abbreviated install document for BOSCO. For the full install document, view Bosco Quick Start Installer.
Bosco Download Visit the Bosco download page. Choose the Quick Start Installer. After downloading the installer, from the terminal, untar it and run the installer as a regular user:
[user@hostname ~]$ tar xvzf ./bosco_quickstart.tar.gz
[user@hostname ~]$ ./bosco_quickstart

Starting Bosco & adding your first cluster using the quick start

login.osgconnect.net is an HTCondor resource I can connect to BOSCO answering the questions:
  • login.osgconnect.net (FQDN of the cluster) 
  • myuser (my user name) 
  • condor (queue manager in the cluster) 

Run the code 

You can test the job locally on your laptop/workstation or login.osgconnect.net but then all the jobs will run on different resources across OSG, all with different software stacks. So I don't know if the software I'm using will be available.
Anyway there are software packages like R and Anaconda (Python, NumPy, ...) available across almost all of OSG thanks to PALMS and OASIS. And I can make sure that my job runs on a node where PALMS and OASIS are available. See Software access using OASIS for detailed instructions on how to use PALMS and OASIS and for a list of the available software.
Python 2.7.5 with all the scientific libraries (the ones included in Anaconda) is all that I will need for my jobs.
Here is the python script counting the "cells" in the image. This is a simplified version. On a real analysys the treshold is more complex than a simple average and some pre-filtering of the image or pattern recognition are used to improve the results:
#!/bin/env python
# count_cells.py - script counting the cells in the image

import numpy as np
from scipy import ndimage
import Image
from scipy.misc import fromimage

# Import image
pic = Image.open("./cellimage.tiff")
im = fromimage(pic)

# Evaluate treshold and segment the image
T = im.mean()
mask = im > T
label_im, nb_labels = ndimage.label(mask)

# print result
print nb_labels
Here is a wrapper script that downloads the image file, sets up the environment and launches the python script. Replace USER_NAME with your username on login.osgconnect.net:
#!/bin/sh
# job_wrapper.sh - script guiding the job execution
IMAGE_FILE=cellimage.tiff
IMAGE_URL=http://stash.osgconnect.net/+USER_NAME/images/cellimage$1.tiff

# download the file
wget -O $IMAGE_FILE "$IMAGE_URL" >&2
if [ ! $? -eq 0 ]; then
echo "Wget failed. Trying curl." >&2
curl -o $IMAGE_FILE "$IMAGE_URL" >&2
if [ ! $? -eq 0 ]; then
echo "Also curl failed. Exiting." >&2
ls -l >&2
exit
fi
fi
if [ ! -f ./$IMAGE_FILE ]; then
echo "Missing image file ./$IMAGE_FILE. Exiting." >&2
ls -l >&2
exit
fi

# setup the environment
source /cvmfs/oasis.opensciencegrid.org/osg/palms/setup 1>&2
palmsdosetup anaconda 1>&2

python ./count_cells.py
Here is the HTCondor submit file:
# job.sub - HTCondor submit file
# If submitting from login.osgconnect.net use: universe = vanilla
# When submitting via BOSCO use:
universe = grid

# remember to create the batch1 directory
Error = batch1/job.err.$(Cluster)-$(Process)
Output = batch1/job.out.$(Cluster)-$(Process)
Log = batch1/job.log.$(Cluster)

# Remember to set the job_wrapper.sh es executable (chmod +x):
Executable = job_wrapper.sh
Arguments = $(Process)
Transfer_input_files = count_cells.py
ShouldTransferFiles = YES
When_to_transfer_output = ON_EXIT

Requirements = (HAS_CVMFS_oasis_opensciencegrid_org =?= TRUE) && (CVMFS_oasis_opensciencegrid_org_REVISION >= 729)
+ProjectName = "ConnectTrain"

# Submitting 100 jobs to analyze 100 different images
queue 100
To submit the jobs I normal HTCondor commands as documented in the BOSCO quickstart. condor_submit returns a job ID that you will then use for the following commands (referred as JOB_ID below). To be more precise the number returned is the job cluster ID, then within my cluster there are 100 job processes (0 to 99), each analyzing a different image:
  • to submit: condor_submit job.sub 
  • then to monitor the jobs running: watch -n 2 condor_q JOB_ID 

Once the jobs completed, the result (cells counted in the image) was returned in the jobs standard output that is copied locally. To sum all the cells I evaluated the total. Replace jour job ID (returned by condor_submit) in the command below:
cat batch1/job.out.JOB_ID-* | paste -sd+ - | bc

References and additional material

Here the links to some more proper cell counting articles. You can take any code and adapt your workflow to run in parallel on OSG:



News and Announcements from OSG Operations > Be a mentor at PEARC17 and make a difference!

Make a difference in the lives of next-generation scientists and science leaders at PEARC17—Practice & Experience in Advanced Research Computing 2017 in New Orleans, July 9-13. As a mentor, you will give students the opportunity to learn from your expertise as they explore a variety of related issues, such as differing specialties, postgraduate training, and work/life balance; plus enjoy Monday’s complimentary student-mentor dinner. Many mentorships turn into lifetime friendships and have countless benefits.  

Check "I want to be a mentor" when you register for the PEARC17 conference, which will take place in New Orleans, July 9-13. The conference is open to professionals and students in advanced research computing. Registration for the conference and hotel are now open.


ABOUT PEARC17

Practice & Experience in Advanced Research Computing 2017—unites the high-performance computing and advanced digital research communities, addressing the challenges of using and operating advanced research computing within academic and open science communities. Being held in New Orleans July 9-13, PEARC17 offers a robust technical program, as well as networking, professional growth and multiple student participation opportunities. See pearc17.pearc.org for more information, and follow us PEARC on Twitter at (PEARC_17) and on Facebook (PEARChpc).

News and Announcements from OSG Operations > Center for Network and Storage-Enabled Collaborative Computational Science Symposium

Dear Colleagues,

Please join us May 18-19 in Ann Arbor for the first symposium of the Center for Network and Storage-Enabled Collaborative Computational Science (CNSECCS).

The symposium is intended to bring together scientists, researchers and infrastructure specialists to share experiences and best practices, and to discuss both challenges and possible solutions that enable scientists to work together on “big, distributed or diverse data”.

Speakers will come from a wide range of research domains, as well as federal funding agencies:

  * Amy Friedlander, Deputy Division Director, Division of Advanced Cyberinfrastructure (CISE/ACI), National Science Foundation.
  * Richard Carlson, Program Officer, Advanced Scientific Computing Research (ASCR), Department of Energy
  * Nina Silverberg, Program Director, Alzheimer's Disease Centers program, Division of Neuroscience, National Institutes of Health

We are also accepting abstracts from potential presenters through the beginning of April.

For more information and to register, visit:
http://micde.umich.edu/centers/cnseccs/2017-symposium/

For more about the center, see:
http://micde.umich.edu/centers/cnseccs/

And for any questions, please email
micde-contact@umich.edu.

We hope to see you there!

Best regards,

Shawn McKee, Director, CNSECCS

News and Announcements from OSG Operations > XRootD 4.6.0 removed from OSG 3.3.22

XRootD 4.6.0 was pulled from the OSG repositories. The OSG repositories still contain XRootD 4.5.0.
XRootD 4.6.0 was in the OSG repositories from Tuesday, March 11th until Friday, March 17th.

End users encountered XRootD failures when using GSI authentication.
The problems have been traced to a defect in the CRL verification code.
XRootD running in a server mode will likely experience problems,
since reading the served file usually requires authentication.
XRootD running in the manager mode may not experience any difficulties,
since the redirector typically does not do authentication.

To check if you have the affected components installed, run
`rpm -qa | grep xrootd` to display the versions of your xrootd packages.
If any of them are version 4.6.0, run `yum downgrade <package>` on those
packages.

We are awaiting the 4.6.1 XRootD release which corrects these issues.
We will test and release it next available OSG release.

We apologize for any inconvenience. Our release testing did not catch
this problem. We are augmenting our release processes to catch this
specific problem as well as similar issues that may come up in other
packages.

News and Announcements from OSG Operations > GOC Service Update - Tuesday, March 28th at 14:00 UTC

The GOC will upgrade the following services beginning Tuesday, March 28th at 14:00 UTC. The GOC reserves 8 hours in the unlikely event unexpected problems are encountered.

Condor Collectors
- Update HTCondor View

Data
- Update internal database servers data1 and data 2

Repo
- Rebuild production instances at RHEL6

Web Services
- Apply standard software updates

All services
- Operating system updates; reboots will be required. The usual HA mechanisms will be used, but some services will experience brief outages.

News and Announcements from OSG Operations > OSG BDII Decommissioning

OSG Collaborators,

As has been previously announced, OSG Operations and Technology are planning the retirement of the BDII information service located at is.grid.iu.edu on March 31st, 2017. As of that date, the SLA for the BDII will expire. In addition to this, on April 4th, at 9:00 US Eastern time, Operations will power off both instances of the BDII.

This is the final announcement regarding the decommissioning of this service.

Pegasus news feed > Pegasus Featured in Blue Waters Webinar Series

Blue Waters provides a series of free webinars throughout the year. The Scientific Workflows webinar track provides an overview of common scientific workflows and tools that enable them. In the first presentation of series of webinars, Scott Callaghan presented an overview of scientific workflows, where he discussed what the community means by “workflows” and what elements make up a workflow.

The presentation features Pegasus WMS as the solution adopted by the Southern California Earthquake Center (SCEC) to address challenges of large scale scientific computations in the Cybershake workflow, which include workflow automation, data management, resource provisioning (mover 500+K jobs through the cluster efficiently), and error handling.

The video for the presentation is shown below, and the full description of the webinar is available here. This presentation has also been featured in insideHPC.

 

The Pegasus team will also present a webinar in the Blue Waters Scientific Workflow webinar track:

Blue Waters Webinar: Pegasus Workflow Management System
Link: https://bluewaters.ncsa.illinois.edu/webinars/workflows/pegasus
Date: May 24, 2017
Presenter: Mats Rynge

 

 

 

 

82 views


News and Announcements from OSG Operations > Announcing OSG Software version 3.3.22

We are pleased to announce OSG Software version 3.3.22.

Changes to OSG 3.3.22 include:
- HTCondor: Patched to avoid Job Router crash when submit requirements fail
- New default configuration: GUMS uses proper TLS version(s) with VOMS admin
- BLAHP 1.18.29: Better Slurm integration, fixed problem with proxy refresh
- HTCondor-CE 2.1.4: Respect 'RequestCpus', Added JSON attributes for AGIS
- Gratia probe 1.17.4: Now picks up 'RequestCpus' with HTCondor-CE
- Update to CVMFS 2.3.3
- Update to GlideinWMS 3.2.18
- Update to XRootD 4.6.0
- HDFS: GridFTP prints proper error message when HDFS quota is exhausted
- VOMS 2.0.14-1.3: Now validates top-level group of proxy

Changes to the Upcoming Repository include:
- HTCondor 8.6.1: New stable series of HTCondor in Upcoming
- Update to frontier-squid 3.5.24

Release notes and pointers to more documentation can be found at:

https://www.opensciencegrid.org/bin/view/Documentation/Release3/Release3322

Need help? Let us know:

https://www.opensciencegrid.org/bin/view/Documentation/Release3/HelpProcedure

We welcome feedback on this release!

Pegasus news feed > Pegasus supports improved delivery of RNA drugs

Water is seen as small red and white molecules on large nanodiamond spheres. The colored tRNA can be seen on the nanodiamond surface. Image by Michael Mattheson, OLCF, ORNL (Source: ORNL)

The Pegasus team is very happy to hear about the cutting-edge research and development that can help solve some of the challenges associated with drug delivery discovery: “The motions of a tRNA (or transfer RNA) model system can be enhanced when coupled with nanodiamonds, or diamond nanoparticles approximately 5 to 10 nanometers in size”

 

Diamonds that deliver!

Neutrons, simulation analysis of tRNA-nanodiamond combo could transform drug delivery design principles.

A unique combination of experimentation and simulation was used to shed light on the design principles for improved delivery of RNA drugs, which are promising candidates in the treatment of a number of medical conditions including cancers and genetic disorders.

Nanodiamonds are good delivery candidates due to their spherical shape, biocompatibility and low toxicity. And because their surfaces can be easily tailored to facilitate the attachment of various medicinal molecules, nanodiamonds have tremendous potential for the delivery of a vast range of therapies.

The discovery involved ORNL’s Spallation Neutron Source, which provides the most intense pulsed neutron beams in the world for scientific research and industrial development, and ORNL’s Titan supercomputer, the nation’s most powerful for open science. By comparing the SNS neutron scattering data with the data from the team’s molecular dynamics simulations on Titan, the researchers have confirmed that nanodiamonds enhance the dynamics of tRNA when in the presence of water; This analysis was performed with the Pegasus WMS.

The DOE Panorama project has developed an SNS workflow to confirm that nanodiamonds enhance the dynamics of tRNA in presence of water. The workflow, enacted by the Pegasus Workflow Management System, calculates the epsilon that best matches experimental data. These calculations were for 10 ns each and the workflows used almost 400,000 CPU hours of time on DOE leadership class systems.

The Panorama project delivers workflow management capabilities, models the end-to-end performance of the workflows on a variety of computational platforms, provides monitoring capabilities for the workflows and the computational infrastructure at runtime, and detects anomalies in the execution when application performance drops below expectations

(Funding Acknowledgements: DOE DE-SC0012636)

 

Read the full press release @ORNL... #kadbtn43:hover {background:#99cccc !important;color:#1c1c1c !important;border-color:#595959!important;}

 

Additional Information:

 

119 views


Miha's Blog > A new HPC startup: Exabyte.io

One of the recurring questions in High Performance Computing is why very little effort goes into users experience. People assume that scientists surely know how to use SuperComputers and clusters and grid computing. Actually, the great majority of scientists are SuperComputing hands-on illiterate

The article Comments to the history of grid describes how scientists are frustrated by the lack of ease of use of their application requiring High Throughput Distributed Computing and /or  High Performance Computing,

A big surprise occurred when Microsoft acquired and will integrate GreenButton - smart, yet small HPC company from New Zealand - into Azure to render movies, democratizing the industry.

Now I discovered Timur Bazhirov, the founder of a new company in San Francisco, exabyte.io

Timur, is a scientist himself, he has PH.D. in Physics from UC. Berkeley, with a decade long experience in computational materials science and molecular modeling using supercomputing facilities.

He noticed the HPC user interface is outdated and that the entry barriers to HPC are high. Secure shell command line is commonly the only available tool and the existing graphical user interfaces are slow and non-intuitive.

Timur states the obvious for anyone working in HPC: the high entry barriers limit the number of HPC users

He created a web based platform with on-demand access to HPC applications. He designed (beta stage) an intuitive user interface collaboration tools, learning aids, data aggregation.



Note: This blog and its' author, Miha Ahronovitz have no association whatsoever with exabyte.io and does not endorse that company in any way.

News and Announcements from OSG Operations > PEARC17 paper submission deadline extended to March 13

PEARC17 to be held July 9-13, 2017, in New Orleans, Louisiana

The technical paper submission (Technology; Software & Data; Workforce, Diversity, and Evaluation; and Accelerating Discovery)  deadline for the inaugural PEARC Conference has been extended one week to March 13, 2017. Technical papers must be submitted through the PEARC17 EasyChair site at https://easychair.org/my/conference.cgi?conf=pearc17.

The full PEARC17 Call for Participation contains details about each of the four technical tracks of papers. The technical track paper submissions may be full papers (strongly preferred) or extended abstracts.

PEARC17, which will take place in New Orleans, July 9-13, 2017, is open to professionals and students in advanced research computing. Registration for the conference and the hotel are now open. Poster, Visualization Showcase and Birds-of-a-Feather submissions are due by May 1. Opportunities for student travel support to the conference are available via the PEARC17 Student Program.

The PEARC (Practice & Experience in Advanced Research Computing) conference series is being ushered in with support from many organizations and will build upon earlier conferences’ success and core audiences to serve the broader community. Organizations supporting the new conference include XSEDE, the Advancing Research Computing on Campuses: Best Practices Workshop (ARCC), the Science Gateways Community Institute (SGCI), the Campus Research Computing Consortium (CaRC), the ACI-REF consortium, the Blue Waters project, ESnet, Open Science Grid, Compute Canada, the EGI Foundation, the Coalition for Academic Scientific Computation (CASC), and Internet2.

Follow PEARC on Twitter (PEARC_17) and on Facebook (PEARChpc).

PEARC17—Practice & Experience in Advanced Research Computing 2017—is building on the successful XSEDE conference series and expanding its scope to include everyone engaged with the challenges of using and operating advanced research computing on campuses or for the academic and open science communities. Being held in New Orleans July 9-13, PEARC17 offers a robust technical program, as well as networking, professional growth and multiple student participation opportunities. See pearc17.pearc.org for details.

News and Announcements from OSG Operations > Announcing OSG CA Certificate and VO Package Updates

Please announce this release at your earliest convenience.

We are pleased to announce a data release for the OSG Software Stack.
Data releases do not contain any software changes.

This release contains updated CA Certificates based on IGTF 1.81:
- Added accredited DarkMatter classic QV-intermediate ICAs (AE)
including QuoVadis Root CA 2 G3 and Root CA 3 G3 higher level CAs (BM)
- Updated contact information for EUN EG-GRID CA (EG)
- Withdrawn classic UKeScienceCA-2A in advance of repurposing (UK)

This release also contains VO Package v71:
- Removed INFN CDF VOMS servers

Release notes and pointers to more documentation can be found at:

https://www.opensciencegrid.org/bin/view/Documentation/Release3/Release33213

Need help? Let us know:

https://www.opensciencegrid.org/bin/view/Documentation/Release3/HelpProcedure

We welcome feedback on this release!

Condor Project News > HTCondor 8.7.0 released! ( March 2, 2017 )

The HTCondor team is pleased to announce the release of HTCondor 8.7.0. This development series release contains new features that are under development. This release contains all of the bug fixes from the 8.6.1 stable release. Enhancements in the release include: Performance improvements in collector's ingestion of ClassAds; Added collector attributes to report query times and forks; Removed extra white space around parentheses when unparsing ClassAds; Technology preview of the HTCondor Annex. Further details can be found in the Development Version History and the Stable Version History. HTCondor 8.7.0 binaries and source code are available from our Downloads page.

Condor Project News > HTCondor 8.6.1 released! ( March 2, 2017 )

The HTCondor team is pleased to announce the release of HTCondor 8.6.1. Highlights of this release are: condor_q works in situations where user authentication is not configured; Updates to work with Docker version 1.13; Fixed several problems with the Job Router; Update scripts to support current versions of Open MPI and MPICH2; Fixed a bug that could corrupt the job queue log when the disk is full. More details about the fixes can be found in the Version History. HTCondor 8.6.1 binaries and source code are available from our Downloads page.

News and Announcements from OSG Operations > First major PEARC17 Conference call-for-participation deadline approaching

First major PEARC17 Conference call-for-participation deadline approaching
 
Technical papers and tutorials proposals for the inaugural PEARC conference for advanced research computing due March 6
 
The first major call-for-participation deadline for the inaugural PEARC Conference is March 6, 2017. PEARC17, which will take place in New Orleans, July 9-13, 2017, is open to professionals and students in advanced research computing. Technical papers and tutorials submissions are due by March 6 and must be submitted through EasyChair, which can be found on the PEARC17 Call for Participation webpage here.

The official Call for Participation contains details about each of the four technical tracks of papers, and tutorials. The technical track paper submissions may be full papers (strongly preferred) or extended abstracts. External Program and Workshop proposals are due March 31 and Poster, Visualization Showcase and Birds-of-a-Feather submissions are due May 1.

The PEARC (Practice & Experience in Advanced Research Computing) conference series is being ushered in with support from many organizations and will build upon earlier conferences’ success and core audiences to serve the broader community. In addition to XSEDE, organizations supporting the new conference include the Advancing Research Computing on Campuses: Best Practices Workshop (ARCC), the Science Gateways Community Institute (SGCI), the Campus Research Computing Consortium (CaRC), the ACI-REF consortium, the Blue Waters project, ESnet, Open Science Grid, Compute Canada, the EGI Foundation, the Coalition for Academic Scientific Computation (CASC), and Internet2.

Follow PEARC on Twitter (PEARC_17) and on Facebook (PEARChpc).

Pegasus news feed > Pegasus 4.7.4 Released

We are pleased to announce Pegasus 4.7.4. Pegasus 4.7.4 is a minor release of Pegasus and includes improvements and bug fixes to the 4.7.4 release. Following bugs were fixed

  1. [PM-1148] – kickstart should print a more helpful error message if the executable is missing
  2. [PM-1160] – Dashboard is not recording the hostname correctly
  3. [PM-1163] – Confusing error message in pegasus-kickstart
  4. [PM-1164] – worker package in submit directory gets deleted during workflow run

 

163 views


Condor Project News > HTCondor Week 2017 registration open ( February 23, 2017 )

Registration is now open for HTCondor Week 2017. The registration deadline is April 17, 2017.

News and Announcements from OSG Operations > GOC Service Update - Tuesday, February 28th at 14:00

The GOC will upgrade the following services beginning Tuesday, 28/Feb at 14:00 UTC. The GOC reserves 8 hours in the unlikely event unexpected problems are encountered.

Data

* Update internal database servers data1 and data 2

Reports

* Upgrade production report generator to CentOS 6.8

VOMS

* Upgrade production instance to CentOS 6.8
* Take over operation of OSG VO voms from FNAL

Twiki

* Back up configuration changes

Web Services

* Basic updates and patches to OSG web presence.

All Services

* Operating system updates; reboots will be required. The usual HA mechanisms will be used, but some services will experience brief outages.

News and Announcements from OSG Operations > Announcing OSG CA Certificate Update

We are pleased to announce a data release for the OSG Software Stack.
Data releases do not contain any software changes.

This release contains updated CA Certificates based on IGTF 1.80:
- Discontinued BEGrid2008 (BELNET) classic authority (BE)

Release notes and pointers to more documentation can be found at:

https://www.opensciencegrid.org/bin/view/Documentation/Release3/Release33212

Need help? Let us know:

https://www.opensciencegrid.org/bin/view/Documentation/Release3/HelpProcedure

We welcome feedback on this release!

Condor Project News > HTCondor Week 2017 web page available ( February 15, 2017 )

The HTCondor Week 2017 web page is now available. This web page includes information about nearby hotel options (note that HTCondor Week 2017 will be held at a different location than the last few HTCondor Weeks, so that may affect your hotel choice). Registration should be open by the end of February; at this time we anticipate a registration fee of $85/day.

News and Announcements from OSG Operations > Announcing OSG Software version 3.3.21

We are pleased to announce OSG Software version 3.3.21.

Changes to OSG 3.3.21 include:
- osg-configure 1.6.1: Additional support for ATLAS AGIS
- GlideinWMS 3.2.17: See Release Notes
- Important bug fixes for CVMFS server
- HTCondor 8.4.11: Final bug fix release of the 8.4 series
- HTCondor-CE 2.1.2: Avoid crash by being more liberal in what we accept
- Added 2 scripts to manage VO and CA certificates in tarball installations

Changes to the Upcoming Repository include:
- Singularity 2.2.1: Moderate Severity security update

Release notes and pointers to more documentation can be found at:

https://www.opensciencegrid.org/bin/view/Documentation/Release3/Release3321

Need help? Let us know:

https://www.opensciencegrid.org/bin/view/Documentation/Release3/HelpProcedure

We welcome feedback on this release!


Subscribe