How are tapes labeled?
I don't want to think about which tape to use.
With date of last backup as well as a volume-name and an increasing index number (e.g., user1, admin1, gis1, user2, admin2, gis2, user3, admin3, gis3, user4, etc.).
How often are tapes rotated in/out?
Oldest tape every 12 weeks, before that week's backup. Check out the tape backup/rotation schedule.
How often do we backup?
Every two weeks.
On what schedule are backups run?
M, T, W via cron w/ e-mail notification of fail/success and reminder to change tape. After all 3 backups have run in a week (as appropriate), send e-mail reminder to clean the tape drive (or setup cron task).
How many tapes in rotation?
Six (6) for each partition.
When do we test our backups?
Every 12 weeks.
How do we test our backups?
How often do we clean the tape drive?
Cleaning tape says it is drive-dependent, suggested frequency for our drive is once a month (assumed usage?)
Clean tape drive every 18 uses. Should map to every 6 weeks, assuming we backup 3 volumes (one volume per tape) every 2 weeks.
Where do we store the archived tapes?
Maximum times a tape can be used?
Manufacturer says certified 260 times, we'll say 200.
How does the number of times a tape has been used for backup affect that tape's life expectancy?
After some discussion, James and I built a spreadsheet (OpenOffice) outlining our tape backup/rotation schedule: Tape Rotation Schedule (ODS 21KB). The starting "week of" date (upper left) can be changed if we don't want to start counting backup date from the week of 4/13/2009.
The tape rotation schedule is also available as a PDF: Tape Rotation Schedule (PDF 47KB)
We backup the user partition every other Tuesday morning (early AM) and the admin partition every other Wednesday morning. It is the responsibility of whoever is on call to change the tapes.
When it is your turn to swap the tapes, please do the following:
Network Access Configuration
| cbi | 165.95.1.128 | 255.255.255.128 |
| tamucc | 165.95.0.0 | 255.255.240.0 |
Note: Openfiler uses conary for package management. To update: conary updateall
In /opt/openfiler/etc/httpd/conf/httpd.conf, changed Listen 446 to Listen 444 (apparently 446 is blocked)
/u3/pharosdb/var/smcorr.dat