Silo Migration - Frequently Asked Questions
Question 1
Question: What file systems are being affected by this change?
Answer: Files located in $ARCHIVE and /projects reside on file systems involved in this migration. Files residing in $HOME, $CENTER and /center/d will not be affected by this change.
Question 2
Question: What can I do to assist with the tape migration?
Answer: There are several things you can do to assist with this migration:
- If you know that you need to use particular data in the next 3 months, please issue a batch_stage on that data. This will ensure the data is brought online and will help us determine the order of data migration.
- Review any data you have in $ARCHIVE and /projects and remove any data that is no longer needed.
- Please be patient if you experience longer than normal stages after the migration has begun.
- If you have data residing in $CENTER which you plan to copy to $ARCHIVE or /projects, please start copying the data to the appropriate file system as soon as possible.
Question 3
Question: What is ARSC doing to reduce the impact of the tape migration?
Answer: We are doing several things:
- We have been carefully reviewing data access patterns to attempt to bring recently used data online during the migration.
- We are increasing the size of the online $ARCHIVE disk pool to nearly 400TB during this migration.
- We are increasing the maximum file size for files saved to the "Copy 1" disk pool. This will allow more small files to be quickly brought back online without access a tape drive.
Question 4
Question: I have a significant number of older files which I need to use but wasn't able to stage prior to 1/28. What is the best way to bring these files online?
Answer: The "batch_stage" command is generally the best way to stage files. If you find that a significant number of files have not yet been staged, contact the ARSC Help Desk to request the particular files and/or directories be expedited, however running batch_stage prior to 1/28 is the best solution.
