Proactive suggestions

[awnews]

Three proactive suggestions:

 

1) An option to delay the start of proactive backup for a specified period (e.g. N minutes) after a proactive client becomes visible. This would make sure that the proactive backup doesn't "pounce" on users as soon as they turn on their machine (yes, I know they can defer, *if* they notice and understand and hit the button) by trying to work a proactive backup into a daily routine. I find that many users use their machines right after they turn them on (email, surfing, printing) and then use them sporadically thru the day, so delaying the backup for a while would have impact on users.

 

2) An option to add a (randomization?) delay between proactive backups. Right now, if 10 proactive backups are queued, they all run immediately and sequentially (e.g. after many computers are turned on in the morning at 8AM). The computers being used are of course impacted but so is the network as there's a large "spike" for a period of time as the backups are run sequentially. The network users wouldn't feel as much of an impact if the (proactive) backups were spread out more evenly over the day. An example might be to allow a global setting of "wait from [N] to [M] minutes between subsequent proactive backups."

 

3) A bandwidth limit (not the same as the Client User--Backup priority) to control how much network bandwidth (% of link? N MB/sec average?) are used by each backup, all backups, etc. This would avoid having Retrospect hog network bandwidth while doing backups at the expense of slowing down each backup. It may be helpful to allow this to be set per script since the bandwidth limit may depend on time of day, which machines (fast or slow, connected to a 10 or 100 or 1000M LAN) are being backed up, etc.