We are using BE10D, planning to upgrade to 12 as soon as our upgrade from Exchange 5.5 is done.. We have an Overland NEO2000 library with two HP LTO3 drives. The system has been flawless, however, I'm noticing the compression ratio on all of our jobs is dropping.. We've had the system since Oct '06, and have never been prompted to run a cleaning tape. We've contacted HP and Overland, and they both state that the LTO series of tape drives should only be cleaned when the drive asks for it.. Trouble is, we've never been asked for a cleaning in over two years of use. When we started with the new system in Oct 06, we regularly got 1.7-1.8:1 compression on our ~6TB/wk of data. Just this past weekend, on our Monthend full job, with one server containing about 1.5TB of data offline, the backup set was reduced by that amount, and I'm seeing an average of 1.125:1 on this. Where we would normally require 7 tapes for the weekend full set, I gave the library 8 tapes and I'm getting prompted for a 9th.. The dataset being backed up has not changed significantly over the two years, and the ratio has been very slowly but steadily dropping since we started with the Overland system, such that now, we're nearly down to the native capacity of the drive. My question is this: Could the fact that the drives have never been cleaned be causing this? The soft error count on the tapes/drives are quite low, especially as viewed by someone who got most of his experience with the older DLT/SDLT systems. We have a couple of tapes with a bit over a hundred soft r/w errors, and the rest are in the low two-digits.. We were told to watch the soft error counts when the drives began to get close to needing a cleaning.. Anybody got any input on this? looking for answers...
Dave Frandin
