| 2017-01-13 08:46:19 |
Andreas Scheuring |
bug |
|
|
added bug |
| 2017-01-13 08:52:34 |
Andreas Scheuring |
description |
DPM is using shared IFL processors. DPM was configured to use 2 processors:
nova hypervisor-show foo
+---------------------------+----------------+
| Property | Value |
+---------------------------+----------------+
| cpu_info_arch | s390x |
| cpu_info_features | [] |
| cpu_info_model | host |
| cpu_info_topology_cores | 2 |
| cpu_info_topology_sockets | 0 |
| cpu_info_topology_threads | 1 |
| cpu_info_vendor | IBM |
| vcpus | 2 |
| vcpus_used | 0 |
+---------------------------+----------------+
Now if you start an instance with 2 processors the output looks like this:
nova hypervisor-show foo
+---------------------------+----------------+
| Property | Value |
+---------------------------+----------------+
| cpu_info_arch | s390x |
| cpu_info_features | [] |
| cpu_info_model | host |
| cpu_info_topology_cores | 2 |
| cpu_info_topology_sockets | 0 |
| cpu_info_topology_threads | 1 |
| cpu_info_vendor | IBM |
| vcpus | 2 |
| vcpus_used | 2 |
+---------------------------+----------------+
If you start another instance with 2 processors the output still looks like this:
nova hypervisor-show foo
+---------------------------+----------------+
| Property | Value |
+---------------------------+----------------+
| cpu_info_arch | s390x |
| cpu_info_features | [] |
| cpu_info_model | host |
| cpu_info_topology_cores | 2 |
| cpu_info_topology_sockets | 0 |
| cpu_info_topology_threads | 1 |
| cpu_info_vendor | IBM |
| vcpus | 2 |
| vcpus_used | 2 |
+---------------------------+----------------+
--> That's confusing! Should we reconsider the the way of how vcpus are calculated?
Is there a factor how many times a IFL processor can be shared? If so, maybe vcpu should be
factor * IFLs
Thoughts? |
DPM is using shared IFL processors. DPM was configured to use 2 processors:
nova hypervisor-show foo
+---------------------------+----------------+
| Property | Value |
+---------------------------+----------------+
| cpu_info_arch | s390x |
| cpu_info_features | [] |
| cpu_info_model | host |
| cpu_info_topology_cores | 2 |
| cpu_info_topology_sockets | 0 |
| cpu_info_topology_threads | 1 |
| cpu_info_vendor | IBM |
| vcpus | 2 |
| vcpus_used | 0 |
+---------------------------+----------------+
Now if you start an instance with 2 processors the output looks like this:
nova hypervisor-show foo
+---------------------------+----------------+
| Property | Value |
+---------------------------+----------------+
| cpu_info_arch | s390x |
| cpu_info_features | [] |
| cpu_info_model | host |
| cpu_info_topology_cores | 2 |
| cpu_info_topology_sockets | 0 |
| cpu_info_topology_threads | 1 |
| cpu_info_vendor | IBM |
| vcpus | 2 |
| vcpus_used | 2 |
+---------------------------+----------------+
If you start another instance with 2 processors the output still looks like this:
nova hypervisor-show foo
+---------------------------+----------------+
| Property | Value |
+---------------------------+----------------+
| cpu_info_arch | s390x |
| cpu_info_features | [] |
| cpu_info_model | host |
| cpu_info_topology_cores | 2 |
| cpu_info_topology_sockets | 0 |
| cpu_info_topology_threads | 1 |
| cpu_info_vendor | IBM |
| vcpus | 2 |
| vcpus_used | 2 |
+---------------------------+----------------+
--> That's confusing! Should we reconsider the the way of how vcpus are calculated?
Is there a factor how many times a IFL processor can be shared? If so, maybe vcpu should be
factor * IFLs
On the other hand, if 2 IFLs are available, shall I be able to start an instance with 4 vcpus?
Thoughts? |
|
| 2017-01-24 06:05:53 |
Sreeram Vancheeswaran |
nova-dpm: assignee |
|
Sreeram Vancheeswaran (sreeram-vancheeswaran) |
|
| 2017-01-24 06:05:56 |
Sreeram Vancheeswaran |
nova-dpm: importance |
Undecided |
Medium |
|
| 2017-01-24 06:06:04 |
Sreeram Vancheeswaran |
nova-dpm: status |
New |
Triaged |
|
| 2017-02-14 13:44:43 |
Andreas Scheuring |
summary |
Number available and used vcpus is confusing |
Clarify hypervisor CPU metrics |
|
| 2017-02-14 13:46:29 |
Andreas Scheuring |
description |
DPM is using shared IFL processors. DPM was configured to use 2 processors:
nova hypervisor-show foo
+---------------------------+----------------+
| Property | Value |
+---------------------------+----------------+
| cpu_info_arch | s390x |
| cpu_info_features | [] |
| cpu_info_model | host |
| cpu_info_topology_cores | 2 |
| cpu_info_topology_sockets | 0 |
| cpu_info_topology_threads | 1 |
| cpu_info_vendor | IBM |
| vcpus | 2 |
| vcpus_used | 0 |
+---------------------------+----------------+
Now if you start an instance with 2 processors the output looks like this:
nova hypervisor-show foo
+---------------------------+----------------+
| Property | Value |
+---------------------------+----------------+
| cpu_info_arch | s390x |
| cpu_info_features | [] |
| cpu_info_model | host |
| cpu_info_topology_cores | 2 |
| cpu_info_topology_sockets | 0 |
| cpu_info_topology_threads | 1 |
| cpu_info_vendor | IBM |
| vcpus | 2 |
| vcpus_used | 2 |
+---------------------------+----------------+
If you start another instance with 2 processors the output still looks like this:
nova hypervisor-show foo
+---------------------------+----------------+
| Property | Value |
+---------------------------+----------------+
| cpu_info_arch | s390x |
| cpu_info_features | [] |
| cpu_info_model | host |
| cpu_info_topology_cores | 2 |
| cpu_info_topology_sockets | 0 |
| cpu_info_topology_threads | 1 |
| cpu_info_vendor | IBM |
| vcpus | 2 |
| vcpus_used | 2 |
+---------------------------+----------------+
--> That's confusing! Should we reconsider the the way of how vcpus are calculated?
Is there a factor how many times a IFL processor can be shared? If so, maybe vcpu should be
factor * IFLs
On the other hand, if 2 IFLs are available, shall I be able to start an instance with 4 vcpus?
Thoughts? |
DPM is using shared IFL processors. DPM was configured to use 2 processors:
nova hypervisor-show foo
+---------------------------+----------------+
| Property | Value |
+---------------------------+----------------+
| cpu_info_arch | s390x |
| cpu_info_features | [] |
| cpu_info_model | host |
| cpu_info_topology_cores | 2 |
| cpu_info_topology_sockets | 0 |
| cpu_info_topology_threads | 1 |
| cpu_info_vendor | IBM |
| vcpus | 2 |
| vcpus_used | 0 |
+---------------------------+----------------+
Now if you start an instance with 2 processors the output looks like this:
nova hypervisor-show foo
+---------------------------+----------------+
| Property | Value |
+---------------------------+----------------+
| cpu_info_arch | s390x |
| cpu_info_features | [] |
| cpu_info_model | host |
| cpu_info_topology_cores | 2 |
| cpu_info_topology_sockets | 0 |
| cpu_info_topology_threads | 1 |
| cpu_info_vendor | IBM |
| vcpus | 2 |
| vcpus_used | 2 |
+---------------------------+----------------+
If you start another instance with 2 processors the output still looks like this:
nova hypervisor-show foo
+---------------------------+----------------+
| Property | Value |
+---------------------------+----------------+
| cpu_info_arch | s390x |
| cpu_info_features | [] |
| cpu_info_model | host |
| cpu_info_topology_cores | 2 |
| cpu_info_topology_sockets | 0 |
| cpu_info_topology_threads | 1 |
| cpu_info_vendor | IBM |
| vcpus | 2 |
| vcpus_used | 2 |
+---------------------------+----------------+
--> That's confusing! Should we reconsider the the way of how vcpus are calculated?
The issues probably is, that we have shared CPUs, but each partition can only have be attached to a single shared CPU once. In other words, if you have 3 shared cpus, a partition can have a maximum number of 3 vCPUs.
In addition the other CPU attributes should be clarified as well! (cores, sockets, threads,...) |
|
