Choose timezone
Your profile timezone:
SDCC Liaison Meeting
→
US/Eastern
3-192 (Bldg 510)
3-192
Bldg 510
, ,
Description
Join via BlueJeans (https://bluejeans.com/819381923). Passcode not required. You can also join via phone (Meeting ID: 819 381 923), by calling one of the numbers below:
+1.408.740.7256 (United States)
+1.888.240.2560 (US Toll Free)
+1.408.317.9253 (Alternate number)
Thursday April 9, 2020 Liaison Meeting Minutes
Facility News
- CYBERSECURITY is in discussion with SDCC, working towards a password policy proposal after krb5 www intrusions
Network & Facility Operations
- recent steady pace of on-site work resuming after MINSAFE ended
- progress for purchases and deployments of CSI-dedicated tape library, mover, and JBOD, Lustre MDSes, and GS JBODs
Storage
- JBODs arrived now, targeting next week for deployment
Fabric
- New Supermicro cluster in production for STAR, PHENIX, and ATLAS T1
- Singularity 3.6.1 released bug fix from 3.6.0. After it’s in EPEL we plan to test and then push to the farm
- HTCondor T3 reorganized group mappings to institutions, and now institutions will be marked with a ClassAd
- Discussion over pulling hardware from the pool for a dedicated sPHENIX Jenkins HTCondor backend
General Services
- lookout for announcements on RHEV maintenance, rebooting web servers, ssh gateways, and AFS servers
- user accounts SDCC would like to remove the legacy 8-char limit to username when accounts are requested
- CYBER update on incident with PHENIX web server, moving forward with a 3-phase plan:
- block external access and restore campus access (read/write)
- exract some external facing programs and then restore external access (read-only and no executable scripts)
- improve/update the full-stack of the remaining external facing programs
Tools & Services
- sPHENIX migrating Invenio into production, planning to backup and version control with Gitea
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~`
Topical Discussion: Data Lifecycle Management
NFS home directories (NFS)
- accessible at SDCC from compute and data transfer nodes for personal work area
- designed for and optimized for small to medium files, no high bandwidth or high IOPS to data
- daily backups are kept for 90 days and daily snapshots are kept for 7 days
- small quota/user (~5GB/user), varies by experiment
Project filesystem (GPFS and Lustre)
- accessible at SDCC from compute and data transfer nodes
- targeted primarily at collaborations for high bandwidth, large block access to files
- no snapshots and no backups
- some project filesystems, particularly but not limited to “scratch”, may have policy driven auto-deletion
Software distribution filesystem (OpenAFS and CVMFS)
- read-only, world accessible filesystems
- designed for worldwide, read-only distribution of applications and libraries
Object storage (BNLBox)
- Default quota is 50GB/user
- Write access is limited to authorized users, read-access controlled by data owner
- built-in file versioning and trash bin (30 day retention up to quota limit policy)
- Data in BNLBox Archive folder doesn’t count against user quota and is transparently migrated to HPSS by policy
Cold storage (HPSS tape)
- This service is tailored for large files ( files > 10GB)
- Access to cold storage must be negotiated with SDCC
- HPSS service performance can vary dramatically depending on access and quality of service requirements
- Planning to preserve Archive class of storage for existing experiments
- Exploring transition to Lustre backend to HPSS for transparent tape backups from the user
There are minutes attached to this event.
Show them.
