Skip to content

Commit 79e7270

Browse files
ppaluckippalucki
committed
update README, remove out-of-date info
1 parent 1975305 commit 79e7270

File tree

1 file changed

+25
-18
lines changed

1 file changed

+25
-18
lines changed

‎deployment/pcm/README.md

Lines changed: 25 additions & 18 deletions
Original file line numberDiff line numberDiff line change
@@ -85,31 +85,33 @@ More information here: https://kubernetes.io/docs/tutorials/security/ns-level-ps
8585

8686
| Method | Used interfaces | default | Notes |
8787
|---------------|------------------------------------------------------------| -------- | ------------------------------------------------------------------------------------- |
88-
| indirect | perf, resctrl | v | missing energy metrics, requires fix for /pcm/resctrl mount |
89-
| direct | msr | | requires msr module and access to /dev/cpu (non trivial) |
88+
| indirect | perf, resctrl | v | missing energy metrics, |
89+
| direct | msr | | requires msr module and access to /dev/cpu (non trivial) or privileged access |
9090

9191

9292
| Metrics | Available on Hardware | Available through interface | Available through method |
9393
| --------------------- | ----------------------------- | ---------------------------- | ------------------------ |
9494
| core | bare-metal, VM (any) | msr or perf | any |
9595
| uncore (UPI) | bare-metal, VM (all sockets) | msr or perf | any |
9696
| RDT (MBW,L3OCCUP) | bare-metal, VM (all sockets) | msr or resctrl | any |
97-
| energy, temp | bare-metal (only) | msr | msr only! |
97+
| energy, temp | bare-metal (only) | msr | direct |
98+
| perf-topdown | | perf only | indirect |
9899

99100

100-
| Interface | Requirements | Controlled by (env/helm value) | default pcm/helm | Used by source code | Helm Value |
101+
| Interface | Requirements | Controlled by (env/helm value) | default helm | Used by source code | Notes |
101102
|---------------|------------------------------------------------------------|---------------------------------|-----------------------|----------------------------------------------------------|-----------------------------------------------------|
102-
| perf | sys_perf_open() perf_paranoid<=0/privileged/CAP_ADMIN | PCM_NO_PERF | 0 / 1 (!) | programPerfEvent(), PerfVirtualControlRegister() | |
103-
| perf-uncore | sys_perf_open() perf_paranoid<=0/privileged/CAP_ADMIN | PCM_USE_UNCORE_PERF | 0 / 0 | programPerfEvent(), PerfVirtualControlRegister() | |
104-
| perf-topdown | /sys/bus/event_source/devices/cpu/events | | yes | cpucounters.cpp:perfSupportsTopDown() | sysMount (TODO: conflicts with sys/fs) |
105-
| RDT | uses "msr" or "resctrl" interface | PCM_NO_RDT | 0 / 0 | cpucounters.cpp:isRDTDisabled()/QOSMetricAvailable() | PCM_NO_RDT |
106-
| resctrl | RW: /sys/fs/resctrl | PCM_USE_RESCTRL | 0 / 0 | resctrl.cpp | resctrlHostMount |
103+
| perf | sys_perf_open() perf_paranoid<=0/privileged/CAP_ADMIN | PCM_NO_PERF | use perf | programPerfEvent(), PerfVirtualControlRegister() | |
104+
| perf-uncore | sys_perf_open() perf_paranoid<=0/privileged/CAP_ADMIN | PCM_USE_UNCORE_PERF | use perf for uncore | programPerfEvent(), PerfVirtualControlRegister() | |
105+
| perf-topdown | /sys/bus/event_source/devices/cpu/events | sysMount | yes | cpucounters.cpp:perfSupportsTopDown() | TODO: conflicts with sys/fs/resctrl |
106+
| RDT | uses "msr" or "resctrl" interface | PCM_NO_RDT | yes | cpucounters.cpp:isRDTDisabled()/QOSMetricAvailable() | |
107+
| resctrl | RW: /sys/fs/resctrl | PCM_USE_RESCTRL | yes | resctrl.cpp | resctrlHostMount |
107108
| watchdog | RO/RW: /proc/sys/kernel/nmi_watchdog | PCM_KEEP_NMI_WATCHDOG | yes (tries to disable)| src/cpucounters.cpp:disableNMIWatchdog() | |
108-
| msr | RW: /dev/cpu/X/msr + privileged or CAP_ADMIN/CAP_RAWIO | PCM_NO_MSR | 0 / 0 | msr.cpp:MsrHandle() | privileged or values-device-injector.yaml |
109-
| | RW: /dev/mem | ? | 0 / 0 | cpucounters.cpp:initUncoreObjects, pci.cpp:PCIHandleM() | privileged or values-device-injector.yaml |
110-
| | RO/RW: /sys/module/msr/parameters | PCM_NO_MSR | 0 / 0 | msr.cpp:MsrHandle() | sysMount |
111-
| | RW: /proc/bus/pci | PCM_USE_UNCORE_PERF ??? | 0 / 0 | pci.cpp:PCIHandle() | pciMount |
112-
| | RO: /sys/firmware/acpi/tables/MCFG | PCM_USE_UNCORE_PERF ??? | 0 / 0 | pci.cpp:PciHandle::openMcfgTable() | mcfgMount |
109+
| msr | RW: /dev/cpu/X/msr + privileged or CAP_ADMIN/CAP_RAWIO | PCM_NO_MSR | msr is disabled | msr.cpp:MsrHandle() | privileged or some method to access /dev/cpu |
110+
| | RW: /dev/mem | ? | msr is disabled | cpucounters.cpp:initUncoreObjects, pci.cpp:PCIHandleM() | privileged or some method to access /dev/cpu |
111+
| | RO/RW: /sys/module/msr/parameters | PCM_NO_MSR | msr is disabled | msr.cpp:MsrHandle() | sysMount |
112+
| | RW: /proc/bus/pci | PCM_USE_UNCORE_PERF | msr is disabled | pci.cpp:PCIHandle() | pciMount |
113+
| | RO: /sys/firmware/acpi/tables/MCFG | PCM_USE_UNCORE_PERF | msr is disabled | pci.cpp:PciHandle::openMcfgTable() | mcfgMount |
114+
| | energy | | | cpucounters.cpp initEnergyMonitoring() | |
113115

114116
### Validation on local kind cluster
115117

@@ -200,18 +202,20 @@ Deploy with defaults:
200202
helm install pcm .
201203
202204
# Alternatively deploy with NFD and with Prometheus enabled
203-
helm install pcm . --set nfd=true --set podMonitor=true
205+
helm install pcm . --set podMonitor=true
206+
kubectl get podmonitor pcm
207+
helm install pcm . --set nfd=true
204208
205209
# Alternatively deploy with NFD and with Prometheus enabled into own "pcm" namespace
206-
helm install pcm . -n pcm --set nfd=true --set podMonitor=true
210+
helm install pcm . --namespace pcm
207211
```
208212

209213
#### 6) Check metrics
210214

211215
Run proxy in background:
212216
```
213217
kubectl proxy &
214-
# for access from another host TODO to be remove
218+
# for access from another host TODO to be remove (unsecure!!!)
215219
kubectl proxy --address 0.0.0.0 &
216220
```
217221

@@ -222,7 +226,10 @@ kubectl get pods
222226
podname=`kubectl get pod -l app.kubernetes.io/component=pcm-sensor-server -ojsonpath='{.items[0].metadata.name}'`
223227

224228
curl -Ls http://127.0.0.1:8001/api/v1/namespaces/default/pods/$podname/proxy/metrics
225-
curl -Ls http://127.0.0.1:8001/api/v1/namespaces/default/pods/$podname/proxy/metrics | grep DRAM_Writes
229+
curl -Ls http://127.0.0.1:8001/api/v1/namespaces/default/pods/$podname/proxy/metrics | grep L3_Cache_Misses # source: core
230+
curl -Ls http://127.0.0.1:8001/api/v1/namespaces/default/pods/$podname/proxy/metrics | grep DRAM_Writes # source: uncore
231+
curl -Ls http://127.0.0.1:8001/api/v1/namespaces/default/pods/$podname/proxy/metrics | grep Local_Memory_Bandwidth{socket="1",aggregate="socket",source="core"} # source: RDT
232+
curl -Ls http://127.0.0.1:8001/api/v1/namespaces/default/pods/$podname/proxy/metrics | grep DRAM_Joules_Consumed # source: energy
226233
```
227234

228235
or through Prometheus UI/prom tool:

0 commit comments

Comments
 (0)