I dont often downvote because theres this margin between "downvote because its a poor question" and "close it because its retarded" . Where I dont want to do either and would at least like some way of venting my frustration at a question.
@MichaelHampton if it's just formatting that's easy but beyond that it's often too hard to know what the question is about - it's easier to give the OP the opportunity to fix things
I can tolerate a badly written question, where it lacks info theres a comment, but some questions just appear to indicate too little done by the user and I have to fight off the urge to leave a patronizing comment (so I dont comment).
Hm, well sometimes I doubt I'm a good enough judge to determine that, since the standard I have is just a lot higher for 'done enough work'.
I.E I was looking into working out why systemd --user doesnt work in rhel7. "Done enough research" in this case was looking through the source code and seeing they patched it out.
@ewwhite I looked at the f20 source, found the thing that initializes the interface. Compared the same file/line in rhel7, noticed it wasnt there. Grepped all the patches for the line that should be there and found the patch responsible.
The patch filename itself was descriptive. The reason they removed it was not. In true redhat style they refer to a bugzilla bug which nobody can actually access.
A few weeks ago I was looking through the kernel changelogs in relation to a NFS caching performance regression we was having. I grepped out all nfs changes from the kernel rpm changelog, then manually visited all teh bugs (about 150), each and every one of them was private.
only because I want to see those bug reports and I want to be able to submit my patches and have them go somewhere other than bugzilla for months and months on end without ever being looked at.
$ mco facts lsbdistrelease
Report for fact: lsbdistrelease
4.7 found 4 times
4.9 found 2 times
5 found 2 times
5.10 found 8 times
5.3 found 1 times
5.4 found 3 times
5.5 found 3 times
5.6 found 1 times
@DennisKaarsemaker I built a CentOS server a few weeks ago. Rather than have serious bodily injuries caused by frustration from SELinux I just disabled it
A bit like terrorism. There is a lot of bravado but I think it really doesn't change much. Or you end up trading in way too much liberty at the cost of security.
> [11:27 AM]: OBJECTIVELY WHO HAD A BETTER LIFE ELLIOT RODGER OR THE GIRLS HE KILLED. THE ANSWER IS THE GIRLS. SO WHY SHOULD SOCIETY NOT FEEL WORSE FOR ELLIOT
So yeah an entitled rich white kid driving around in a new 3-series totally had a worse life than the people he killed
I am in a hotel in Washington, DC, and my boyfriend is taking a bath, reading. I barge in, demanding to know if all men are terrible, eyes blazing. He tries to calm me down, but I am upset.
Question: When do you guys look at increasing vCPU shares or cores? When a VM regularly sits at >50% CPU? Or when it peaks at 100% for more than $x minutes?
@MarkHenderson When the services seem slow, the CPU utilization is high, and it seems likely that upping one of those things would actually make the services faster.
@MarkHenderson The peaks hitting 90%-100% for more than 5-15 minutes on a regular basis is roughly what I'd mean by "high".
@MarkHenderson You use the the things you know about in your brain and the things you can find out to come up with an idea of how to get from where things are now to where you want things to be. Then partway through you throw all of that away and run around in circles until the fire goes out.
It's almost like you could look for a side-by-side comparison of an Intel X5450 and E5-4620 CPU online and interpret the results!!
For your use case, virtualization is a convenience. You're definitely not maximizing the use of the new platform, but at least you're using a system that is current ...
@ewwhite I don't need any drivers, I just want the SNMP stuff so that my monitoring can see all 42 temp sensors. Some of the stuff from their RHEL6 channel worked fine on RHEL7, but the hp-snmp-agents tries to link against a version of the snmp library that isn't there.
@ewwhite It's an experimental box, not something i'm rolling into production next week. Mostly trying to make sure all my puppet shit is fully ready for RHEL7.
Proposal: Programming for Professionals
