Implements support for arbitrary Performance Co-Pilot
metrics with per-process instance domains to form new
htop columns. The column-to-metric mappings are setup
using configuration files which will be documented via
man pages as part of a follow-up commit.
We provide an initial set of column configurations so
as to provide new capabilities to pcp-htop: including
configs for containers, open fd counts, scheduler run
queue time, tcp/udp bytes/calls sent/recv, delay acct,
virtual machine guests, detailed virtual memory, swap.
Note there is a change to the configuration file path
resolution algorithm introduced for 'dynamic meters'.
First, look in any custom PCP_HTOP_DIR location. Then
iterate, in priority order, users home directory, then
local sysadmins files in /etc/pcp/htop, then readonly
configuration files below /usr/share/pcp/htop. This
final location becomes the preferred place for our own
shipped meter and column files.
The Settings file (htoprc) writing code is updated to
not using the numeric identifier for dynamic columns.
The same strategy used for dynamic meters is used here
where we write Dynamic(name) so the name can be setup
once more at start. Regular (static) columns writing
to htoprc - i.e. numerically indexed - is unchanged.
Currently htop does not support offline CPUs and hot-swapping, e.g. via
echo 0 > /sys/devices/system/cpu/cpu2/online
Split the current single cpuCount variable into activeCPUs and
existingCPUs.
Supersedes: #650
Related: #580
It can happen that pcp-htop is presented multiple definitions
of the same dynamic meter, e.g. if /etc/pcp/htop/meters has a
definition matching one in ~/.config/htop/meters - instead of
exiting with a duplicate metric error provide more meaningful
diagnostics (on close) and also just skip over such entries.
System files override home directories which overrides those
found below the current working directory.
Also fix the derived metric error diagnostic; because this is
using CRT_fatalError, which is like perror(3), we must give a
meaningful prefix (like program name) at the string end.
Several improvements to the way values are displayed in the
PCP platform DynamicMeter implementation:
- handle the initial 'caption' setting as with regular meters,
this required a new meter callback because we no longer have
just a single meter caption for the DynamicMeter case
- if no label is provided for a metric in a configuration file
use the short form metric name as a fallback
- honour the suffix setting in the configuration file
- convert metric values to the canonical units for htop (kbyte
and seconds), and use Meter_humanUnit when it makes sense to
do so.
Also improves the handling of fatal string error messages in a
couple of places, thanks to BenBE for the review feedback.
This commit is based on exploratory work by Sohaib Mohamed.
The end goal is two-fold - to support addition of Meters we
build via configuration files for both the PCP platform and
for scripts ( https://github.com/htop-dev/htop/issues/526 )
Here, we focus on generic code and the PCP support. A new
class DynamicMeter is introduced - it uses the special case
'param' field handling that previously was used only by the
CPUMeter, such that every runtime-configured Meter is given
a unique identifier. Unlike with the CPUMeter this is used
internally only. When reading/writing to htoprc instead of
CPU(N) - where N is an integer param (CPU number) - we use
the string name for each meter. For example, if we have a
configuration for a DynamicMeter for some Redis metrics, we
might read and write "Dynamic(redis)". This identifier is
subsequently matched (back) up to the configuration file so
we're able to re-create arbitrary user configurations.
The PCP platform configuration file format is fairly simple.
We expand configs from several directories, including the
users homedir alongside htoprc (below htop/meters/) and also
/etc/pcp/htop/meters. The format will be described via a
new pcp-htop(5) man page, but its basically ini-style and
each Meter has one or more metric expressions associated, as
well as specifications for labels, color and so on via a dot
separated notation for individual metrics within the Meter.
A few initial sample configuration files are provided below
./pcp/meters that give the general idea. The PCP "derived"
metric specification - see pmRegisterDerived(3) - is used
as the syntax for specifying metrics in PCP DynamicMeters.
Remove code now that we have common platform-independent command
line wrangling (thanks BenBE!). Add PCP platform support for a
handful of other recently arriving odds and ends - ELAPSED time,
CWD, and so on.
Updates for recent NetworkIO Meter changes, adds support
for the SysArch and HostName Meters. The SysArch change
is based on work originally by Sohaib Mohamed.
This introduces an initial platform for extracting metrics
using the PCP performance metrics API - PMAPI(3). It can
be used via the --enable-pcp=yes configure option.
So far I've added support for live localhost metrics only,
and only using pre-defined metrics already found in htop.
If available, all sampling is performed by pmcd(1) - else,
we fallback to htop doing the metric sampling itself (all
below the PMAPI). When pmcd is used, it may be configured
to run children with elevated privileges, so htop does not
need to be setuid (authentication with pmcd is available).
Additionally, the PMAPI allows us to support archives (for
historical analysis and for automated regression tests in
htop). We'll need platform-specific command line argument
additions, which isn't yet feasible in htop (not difficult
to add though).
The goal of this first version is minimal impact in terms
of modifying the htop codebase, to introduce key ideas in
PCP (metric namespace, metadata, APIs and so on) and give
us something to discuss, experiment with and build on.
Sohaib Mohamed