Issues with the archive server persist. We are in contact with the vendor to resolve the issue.

Affected services
1) /illumina/data partition not accessible
2) MySQL database (linga-dbase) not accessible
3) Framingham pipeline scripts that rely on the database

Next update 3/28 @ 2:00 pm

Please note that the LinGA archive server went down on 3/27 @ 7:00 pm

Affected services

• /illumina/data partition not accessible
• MySQL database (linga-dbase) not accessible
• Framingham pipeline scripts that rely on the database

Next update 3/28 @ 10:00 am

Recently we have had a number of questions about storage usage, replication and snapshots. Data backups and integrity are one of the many things we do behind the scenes, but in an effort to be more transparent and answer some of those questions, we have created an informational page with an overview of our primary storage cluster.

Now available in the Documentation section: http://www.bumc.bu.edu/linga/documentation/linga-systems/ibrix-storage-cluster/

We have made available more cores on RubySky, increasing the number of processors from 512 to 1024. If you have been waiting a long time for jobs to run on either RedStar or BlueIce, consider resubmitting these jobs on RubySky.

Thanks,

LinGA Adminsitrators

The LinGA storage system is stable.

Earlier this morning one of the servers from the main LinGA storage cluster crashed and went in failover mode. Soon after its high-availability pair also experienced issues. As a result, our storage system was inaccessible for a short time.

The following partitions were affected:

• /protected/individuals/ (/data/home)
• /protected/projects
• /unprotected/projects (/data/share)

Currently the storage system is back on line but we’ve lost high-availability. Batch running jobs should have picked up where they had left off. Few might have crashed. Please check the logs.

We’re in contact with the vendor and are working on resolving the high-availability issue. As we go through the troubleshooting steps and apply necessary changes the storage might again become temporarily unavailable.

Thank you for your patience and understanding. We’ll keep you posted with the progress.

LinGA’s compute environment is not directly accessible from the internet and accessing the RStudio web-gui requires port-forwarding, something previously undocumented on the LinGA page. Please take a look at the new documentation. The final section is an example for RStudio.

Note that RStudio does not fit LinGA’s shared computing environment and can only be used interactively on the login-node.

VPN – http://www.bumc.bu.edu/linga/documentation/connecting/

PuTTY – http://www.bumc.bu.edu/linga/documentation/connecting/incoming-to-linga/windows/interactive/windows-text-only/

X-Win32 – http://www.bumc.bu.edu/linga/documentation/connecting/incoming-to-linga/windows/interactive/windows-gui/

FTP\FileZilla – http://www.bumc.bu.edu/linga/documentation/connecting/incoming-to-linga/windows/transfer/#Filezilla

Systems were brought online throughout the week and most functionality was returned by Wednesday. The final cluster, RubySky, was enabled for use on Friday. During the outage, we took the opportunity to repair and make cluster-wide configuration changes to the nodes and clusters. Thank you all for your patience.