Look at them for what? If it says verified/successful, and it's about that big... *Shrug* The missing data (files) weren't large enough to raise any flags (maybe 100MB total out of an 85GB backup) ... So it's pretty much a Perfect (oh shit) Storm of events.
-Stoic Joker
If a year's worth of data suddenly drops off your backup you don't notice something is wrong? Or you don't think it's unusual when yesterday's backup had over a million items and today it's around 80% of that? 
-40hz
That's the point I was trying to make, there was no radical change readily apparent. The files in question are relatively tiny, file count and size change constantly (user directories get backed up too), and the client noticed/said nothing. I'm not there on a daily basis watching the goings on, I'm miles away ... So I can only react to what I know. Now if somebody had decided to tell me staff member X will/has/is about to part on shakey terms ... I can make a point of scrutinizing things in more depth. But, That. Didn't. Happen.
Hell the program that created/access the records/files didn't even (errors/warnings/odd behavior) blink ... and it supposed to be SQL based.
You will often need to change how your backup reports are configured since the default is usually either to log nothing - or everything. Unfortunately, most people have them log every detail - which is useless since it can easily run to hundreds of text pages.
But that still doesn't make them totally useless. As you pointed out, most have a file count summary and success/fail report at the end of the job. And I do eyeball for certain especially critical directories. I don't care if the office temp's 'furry porn' collection went missing. But I do worry when accounting or other important directories flag backup errors on half the files.-40hz
I should probably mention that this assumes that the client in question has bothered to invest in a top shelf backup suite that actually generates handy reports with all the details about how it went. As opposed to just giving last run status = 0x0. Oh yeah, and the company is a tax accounting firm.
addendum: another thing worth looking at in the log is the job run times. If backup time is steadily or abruptly increasing without a corresponding increase in the number of items being backed up, that's a red flag. It could indicate hardware issues, excessive file fragmentation, or corruption of the backup system's database or indexes. All are things worth investigating before disaster strikes.
-40hz
Slowly increasing backup time (other options): Dirty tape heads, failing/EOL tape, or (abrupt increase) somebody left QuickBooks (or some other crapware) running again (fun with file locking).
In a perfect world we get these types of detail. However, in SOHO/SMB ... Not so much.
Recent Posts
)
-40hz (July 11, 2011, 03:35 PM)
...I see this happen all too often.