Join Log in
Score:0
Steve Folly avatar Server

Hardware suitability for Kubernetes

cn flag
Steve Folly

I can find many interweb articles about amount of RAM required for Kubernetes, but I'm struggling to find minimum spec CPUs.

I have a machine with Xeon E5410 @ 2.3GHz, and 16GB of RAM.

I'm finding I'm struggling to run k3s on it and deploy just Loki, Grafana and Prometheus (including kube metrics and node exporters) - I see many pod restarts due to readiness probes timing out, and CPU usage is very high - averaging above 50%

I know it's not a 'modern' CPU, but has it really aged that much that it can't cope with a small Kubernetes node?

This is /proc/cpuinfo...

processor       : 0
vendor_id       : GenuineIntel
cpu family      : 6
model           : 23
model name      : Intel(R) Xeon(R) CPU           E5410  @ 2.33GHz
stepping        : 6
microcode       : 0x60f
cpu MHz         : 2327.615
cache size      : 6144 KB
physical id     : 0
siblings        : 4
core id         : 0
cpu cores       : 4
apicid          : 0
initial apicid  : 0
fpu             : yes
fpu_exception   : yes
cpuid level     : 10
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ht tm pbe syscall nx lm constant_tsc arch_perfmon pebs bts rep_good nopl cpuid aperfmperf pni dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr pdcm dca sse4_1 lahf_lm pti tpr_shadow vnmi flexpriority vpid dtherm
vmx flags       : vnmi flexpriority tsc_offset vtpr vapic
bugs            : cpu_meltdown spectre_v1 spectre_v2 spec_store_bypass l1tf mds swapgs itlb_multihit
bogomips        : 4655.23
clflush size    : 64
cache_alignment : 64
address sizes   : 38 bits physical, 48 bits virtual
power management:

processor       : 1
vendor_id       : GenuineIntel
cpu family      : 6
model           : 23
model name      : Intel(R) Xeon(R) CPU           E5410  @ 2.33GHz
stepping        : 6
microcode       : 0x60f
cpu MHz         : 2190.929
cache size      : 6144 KB
physical id     : 0
siblings        : 4
core id         : 1
cpu cores       : 4
apicid          : 1
initial apicid  : 1
fpu             : yes
fpu_exception   : yes
cpuid level     : 10
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ht tm pbe syscall nx lm constant_tsc arch_perfmon pebs bts rep_good nopl cpuid aperfmperf pni dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr pdcm dca sse4_1 lahf_lm pti tpr_shadow vnmi flexpriority vpid dtherm
vmx flags       : vnmi flexpriority tsc_offset vtpr vapic
bugs            : cpu_meltdown spectre_v1 spectre_v2 spec_store_bypass l1tf mds swapgs itlb_multihit
bogomips        : 4655.23
clflush size    : 64
cache_alignment : 64
address sizes   : 38 bits physical, 48 bits virtual
power management:

processor       : 2
vendor_id       : GenuineIntel
cpu family      : 6
model           : 23
model name      : Intel(R) Xeon(R) CPU           E5410  @ 2.33GHz
stepping        : 6
microcode       : 0x60f
cpu MHz         : 2327.615
cache size      : 6144 KB
physical id     : 0
siblings        : 4
core id         : 2
cpu cores       : 4
apicid          : 2
initial apicid  : 2
fpu             : yes
fpu_exception   : yes
cpuid level     : 10
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ht tm pbe syscall nx lm constant_tsc arch_perfmon pebs bts rep_good nopl cpuid aperfmperf pni dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr pdcm dca sse4_1 lahf_lm pti tpr_shadow vnmi flexpriority vpid dtherm
vmx flags       : vnmi flexpriority tsc_offset vtpr vapic
bugs            : cpu_meltdown spectre_v1 spectre_v2 spec_store_bypass l1tf mds swapgs itlb_multihit
bogomips        : 4655.23
clflush size    : 64
cache_alignment : 64
address sizes   : 38 bits physical, 48 bits virtual
power management:

processor       : 3
vendor_id       : GenuineIntel
cpu family      : 6
model           : 23
model name      : Intel(R) Xeon(R) CPU           E5410  @ 2.33GHz
stepping        : 6
microcode       : 0x60f
cpu MHz         : 2327.615
cache size      : 6144 KB
physical id     : 0
siblings        : 4
core id         : 3
cpu cores       : 4
apicid          : 3
initial apicid  : 3
fpu             : yes
fpu_exception   : yes
cpuid level     : 10
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ht tm pbe syscall nx lm constant_tsc arch_perfmon pebs bts rep_good nopl cpuid aperfmperf pni dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr pdcm dca sse4_1 lahf_lm pti tpr_shadow vnmi flexpriority vpid dtherm
vmx flags       : vnmi flexpriority tsc_offset vtpr vapic
bugs            : cpu_meltdown spectre_v1 spectre_v2 spec_store_bypass l1tf mds swapgs itlb_multihit
bogomips        : 4655.23
clflush size    : 64
cache_alignment : 64
address sizes   : 38 bits physical, 48 bits virtual
power management:

What other things should I be looking at to diagnose this please?

Thanks

53
0 + 2
SYN avatar
hk flag
SYN
I'ld investigate the applications individually. Keep in mind vanilla kubernetes (and moreover k3s) can be deployed ro rpi3. While since rpi4, the 4Gi ones have well enough memory and CPU to run prometheus stack, HA,with thanos/grafana/alertmanagers, ... Old CPUs or commodity hardware should not be much of an issue. While regarding loki, I can't say I tried yet ... maybe try to focus on prometheus first, and increase complexity progressively. Either way: figure out why/who is loading (most likely: prometheus+loki, could involve IOPs, make sure you have fast-enough storage)....
0
Reply
SYN avatar
hk flag
SYN
... While for the readiness timing out, you should be able to play on CPU requests, readiness probes timeout intervals .. but readiness won't restart pods: also check for the liveness probes configuration, which might trigger endless pods restarts, which in turn would increase CPU consumption since processes are continuously trying to start up. Consider using startupProbes when it makes sense, with larger failure thresholds/retries, initialDelaySeconds, ...
0
Reply
ธงชาติไทย
Elon Musk
I sit in a Tesla and translated this thread with Ai:

mangohost

Post an answer

Most people don’t grasp that asking a lot of questions unlocks learning and improves interpersonal bonding. In Alison’s studies, for example, though people could accurately recall how many questions had been asked in their conversations, they didn’t intuit the link between questions and liking. Across four studies, in which participants were engaged in conversations themselves or read transcripts of others’ conversations, people tended not to realize that question asking would influence—or had influenced—the level of amity between the conversationalists.