During Murano performance testing was discovered that starting from 22-25 concurrently (it depends on how the load will be distributed among controllers) environment deployments 10-20% deployments failed with error during usage of heat, neutron or keystone clients as Haproxy closes connection with 408 error:
<134>Sep 28 16:40:38 node-126 haproxy[32680]: 172.16.46.57:42623 [28/Sep/2015:16:40:18.037] heat-api heat-api/<NOSRV> -1/-1/-1/-1/20001 408 212 - - cR-- 2/1/0/0/3 0/0 "<BADREQ>"
where
c : the client-side timeout expired while waiting for the client to
send or receive data.
R : the proxy was waiting for a complete, valid REQUEST from the client
(HTTP mode only). Nothing was sent to any server.
2015-09-28 16:40:48.412 13999 DEBUG murano.dsl.executor [-] 604d1bff76404751a7e114876fc2203b: End execution: io.murano.system.HeatStack.updateTemplate (-1777664545764330947) with exception [heatclient.exc.HTTPException]: ERROR: <html><body><h1>408 Request Time-out</h1>
Your browser didn't send a complete request in time.
</body></html>
_invoke_method_implementation /usr/lib/python2.7/dist-packages/murano/dsl/executor.py:156
Apparently, it happens as Murano-engine uses green theads and utilizes only one CPU of controller. And starting from 22-25 concurrently deployments utilize 100% of CPU.
Additionaly about the same percent of failures is caused by Socket closed/Broken pipe errors, apparently also as Murano-engine comes to the state of lack of CPU resources to support heartbeats.
processor : 5
vendor_id : GenuineIntel
cpu family : 6
model : 62
model name : Intel(R) Xeon(R) CPU E5-2620 v2 @ 2.10GHz
stepping : 4
microcode : 0x424
cpu MHz : 1200.000
cache size : 15360 KB
physical id : 0
siblings : 12
core id : 5
cpu cores : 6
apicid : 10
initial apicid : 10
fpu : yes
fpu_exception : yes
cpuid level : 13
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 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc aperfmperf eagerfpu pni pclmulqdq dtes64 monitor ds_cpl vmx smx est tm2 ssse3 cx16 xtpr pdcm pcid dca sse4_1 sse4_2 x2apic popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm ida arat epb xsaveopt pln pts dtherm tpr_shadow vnmi flexpriority ept vpid fsgsbase smep erms
bogomips : 4200.37
clflush size : 64
cache_alignment : 64
address sizes : 46 bits physical, 48 bits virtual
power management:
Going to check a possible workaround: to run 1 additional murano-engine on each of controllers. Will add comment with results.
Upstream bug: https://bugs.launchpad.net/murano/+bug/1504135
Is that the murano engine which utilizes that much of CPU? Not the heat api?