Ask Your Question
0

no valid host error printed due to cpu overcommit [closed]

asked 2017-06-13 08:53:48 -0500

jacobbaek gravatar image

hi everyone,

i failed to make a virtual machine via printing "no valid host".


2017-06-13 13:32:43.889 16912 WARNING nova.scheduler.utils [req-eab4a70b-66be-48ba-909f-e3aa7f056316 0e69bf82b45640afa970912fe75fb1b1 d15f70779c3d45dc8fbea75bb927968f - - -] Failed to compute_task_build_instances: No valid host was found. There are not enough hosts available. Traceback (most recent call last):

File "/usr/lib/python2.7/site-packages/oslo_messaging/rpc/server.py", line 150, in inner return func(args, *kwargs)

File "/usr/lib/python2.7/site-packages/nova/scheduler/manager.py", line 104, in select_destinations dests = self.driver.select_destinations(ctxt, spec_obj)

File "/usr/lib/python2.7/site-packages/nova/scheduler/filter_scheduler.py", line 74, in select_destinations raise exception.NoValidHost(reason=reason)

NoValidHost: No valid host was found. There are not enough hosts available.

2017-06-13 13:32:43.890 16912 WARNING nova.scheduler.utils [req-eab4a70b-66be-48ba-909f-e3aa7f056316 0e69bf82b45640afa970912fe75fb1b1 d15f70779c3d45dc8fbea75bb927968f - - -] [instance: 12723581-4a75-4cb7-a60e-fd09d1238b1b] Setting instance to ERROR state.


i guess this issue would be occurred by cpu overcommit. because

[stack@director9 ~]$ nova hypervisor-stats

| grep vcpu| vcpus | 24 |

| vcpus_used | 18 |

as you can see physical cpu core 24, virtual cpu core 18 as i know virtual cpu core could check by horizon and command line. so it doesn't matter.

by the way, i made 3 virtual machine using 2 socket per vm, it was working. but when i made 4 virtual machine using 2 socket per vm, it doesn't work.

i think my nova.conf is not bad. /etc/nova/nova.conf cpu_allocation_ratio=16.0

i don't know what effect failing to make virtual machine. i used openstack version mitaka. if anyone have some solution, please help me. Thank you.

edit retag flag offensive reopen merge delete

Closed for the following reason the question is answered, right answer was accepted by jacobbaek
close date 2017-06-13 21:41:18.730330

1 answer

Sort by ยป oldest newest most voted
0

answered 2017-06-13 21:40:47 -0500

jacobbaek gravatar image

i found the reason why it is failed. i missed log for failure. the log is below.

2017-06-14 02:37:14.568 1126 DEBUG nova.scheduler.filters.disk_filter [req-eb5864f8-441b-4650-83e7-7ec49110b0b2 0e69bf82b45640afa970912fe75fb1b1 d15f70779c3d45dc8fbea75bb927968f - - -] (overcloud-compute-0.localdomain, overcloud-compute-0.localdomain) ram: 42838MB disk: -24576MB io_ops: 0 instances: 5 does not have 40960 MB usable disk, it only has -24576.0 MB usable disk. host_passes /usr/lib/python2.7/site-packages/nova/scheduler/filters/disk_filter.py:55

it's not related cpu overcommit. it's occurred due to disk size. after changing disk, it is resolved.

edit flag offensive delete link more

Get to know Ask OpenStack

Resources for moderators

Question Tools

1 follower

Stats

Asked: 2017-06-13 08:53:03 -0500

Seen: 211 times

Last updated: Jun 13 '17