| 1 |
On Mon, May 1, 2017 at 2:46 PM, Rich Freeman <rich0@g.o> wrote: |
| 2 |
> On Sun, Apr 30, 2017 at 4:17 PM, Kai Krakow <hurikhan77@×××××.com> wrote: |
| 3 |
>> Am Sun, 30 Apr 2017 10:33:05 -0700 |
| 4 |
>> schrieb Jorge Almeida <jjalmeida@×××××.com>: |
| 5 |
>> |
| 6 |
>>> It makes sense that the kernel has it. Should it be enabled? For a |
| 7 |
>>> server, probably. For a single-user workstation? Maybe. |
| 8 |
>> |
| 9 |
|
| 10 |
> |
| 11 |
> Honestly, I can't think of why you wouldn't want to use it. |
| 12 |
> |
| 13 |
> The use cases of killing orphan processes and managing resources at a |
| 14 |
> service level have already been mentioned. |
| 15 |
|
| 16 |
I don't usually have orphan processes (that process 1 doesn't reap). |
| 17 |
My services don't require fine tuning re resources. |
| 18 |
|
| 19 |
> |
| 20 |
> Another use case is that the kernel automatically takes cgroups into |
| 21 |
> account when scheduling. So, if one of your services launches a bunch |
| 22 |
> of children they'll be weighted together when allocating CPU. That |
| 23 |
> means that a service with ten threads won't get 10x the CPU of a |
| 24 |
> service with one thread if CPU becomes limiting, assuming equal |
| 25 |
> niceness/etc. On a multi-user system the same would apply to the user |
| 26 |
> running 100 processes vs 1. |
| 27 |
> |
| 28 |
> I also use cgroups to monitor memory use/etc at a service level. |
| 29 |
|
| 30 |
I don't have complex services (some might argue that very complex |
| 31 |
services are badly designed services, but I leave that discussion to |
| 32 |
pros). I only run single-user workstations. |
| 33 |
|
| 34 |
> |
| 35 |
> Sure, they're somewhat optional, but they're a pretty useful kernel feature. |
| 36 |
|
| 37 |
No arguing there. Still, it shouldn't be pushed. It's a bad sign. |
| 38 |
|
| 39 |
Jorge |
Archives