|
|
I am online at 6:45 a.m. and notice that of 14 people logged in, seven connections have been inactive for between 2 and 8 hours. Is this a problem for grex, in terms of making things run slower, or otherwise gobbling up resources? If so, is it possible to put something in place that knocks people off the system after, say, an hour or two of inactivity?
6 responses total.
If the person is connected on a network port (ttyp, ttyq, or ttyr) then they're using only negligable resources if they're idle. It would not be worthwhile to bother kicking off idle network users (especially since gregc and I are the two people you'll most often see idle for long periods of time on a network port... :-) ). On the other hand, people idle for long periods (or even logged on for long periods) on the dialin ports (ttyh) are a large drain on a resource, since no one else can use that modem while they're logged on. However, the number of people who stay logged on for a long period of time on a dialin port is fairly small, and so far staff has been okay dealing with them manually. In addition, the task of writing an idle-killer that properly deals with "idle-but-not-idle" connections such as file transfers is fairly large, and probably not worth spending a lot of time on.
Do you consider the possible security problems of someone leaving an idle terminal to be of significance?
And do idle users use up swap space?
Whether an idle terminal would be a security problem would probably depend on where that person was. If they're in a computing center or some other public place, it would be a problem. But if it's in a private office, or in somebody's house, it's probably not a problem. But the thing to keep in mind there is that all anybody could get into was the person's account who left it idle, and couldn't do anything to anybody else that they couldn't do simply by running newuser. So if somebody loses their mail files or their privacy by leaving their computer idle and logged in in a public place, I'd say that's their problem.
Idle users *do* use up swap space. The amount is dependent on what processes are left idling. This can only be determined by examinging output from the ps command. typically, though, I'd guess that these folks aren't leaving more than a csh or sh shell around. I'd expect minimal impact on our total swap space from this.
since we also have a 24 ptty limit right now, each idle user is taking up about 4% of that resource.
Response not possible - You must register and login before posting.
|
|
- Backtalk version 1.3.30 - Copyright 1996-2006, Jan Wolter and Steve Weiss