Open kraj opened 4 years ago
@sitio-couto could you pair with @lcnzg today and check what is going on?
@kraj, I was unable to reproduce the issue, all Debian derivatives seem to work. Are you still having this issue? If so, is there any custom configuration you've used when creating the instance?
I just open ssh port and launch the largest instance nothing special.
@sitio-couto take a look at the placement, maybe one of the P9s is the issue.
@sitio-couto take a look at the placement, maybe one of the P9s is the issue.
I created 4-5 times and same issue all the time. Fedora 29 booted but was very very slow on p9 now I have provisioned a p8 VM and its chugging along ubuntu 19.04 well. I can build yocto images
@sitio-couto take a look at the placement, maybe one of the P9s is the issue.
Both P9's are scheduling the Debian derivatives, although the load is unbalanced and launch times are a bit high (about 5-8 minutes). I'm doing more tests to see if I can get an instance stuck during boot.
@kraj I was able to boot Ubuntu 18.04 and 19.04 on P9.
@rpsene ok. I have exhausted my compute quota for provisioning my p8 vm, for now, once I have done some build tests I will tear it down and re-install with ubuntu 19.04 on p9, How long should I wait for it to boot ?
OK I launched the instance and let it boot, next day I See the console shows the login prompt. ssh also works as seen below. But the instance is so slow its quite visible. Perhaps thats another issue ?
ubuntu@yocto-power9:~$ cat /proc/cpuinfo
processor : 0
cpu : POWER9 (architected), altivec supported
clock : 2250.000000MHz
revision : 2.2 (pvr 004e 1202)
Hey kraj. You're right, the power 9 machines are quite sluggish. We're trying to pinpoint where's the problem, and as soon as we find out and fix it we'll let you know. For now, it might be better to stick with the Power8 availability zone. Apologies for the inconvenience.
@sitio-couto thanks for confirmation. No issues, I am happy to test it out. Let me know when you have it fixed. I started an update on the p9 VM about 6 hrs ago and it has not finished. Usually it should take less than 5 mins.
Hey @kraj. We've done updating and testing or the P9 machines. They seem to be running normally now. Give then a try later and let us know if you have any issues.
I launched a max instance and it boots but I cant ssh into it I see
ssh: connect to host minicloud1 port XXXXX: No route to host
@kraj did you enable ssh in the security group? ssh
@kraj did you enable ssh in the security group? ssh @minicloud.parqtec.unicamp.br -p xxxxx
yes. I did, same works ok when I launch ubuntu 19.04 on ppc8 box.
@kraj I have deployed many VMs on P9 yesterday without any issues. I need to understand the steps you are using and whether or not you as customizing something.
Fact that I launch power8 machine in exact same process kind of validate my approach, I just forcibly chose power9 when trying to get machine type.
Hey @kraj. Besides adding the SSH rule to the default security group, could you double-check the if the following configurations were correct when creating and accessing the instance:
@sitio-couto right. I have exact same steps followed. Booting power9 does take a bit longer. But I also tried to delete/add the ssh rule after system was fully booted but to no avail. I launch max instance and choose power9 during building the VM
In this case, do you remember the VM’s IP or have tried to launch another VM with a different IP?
My current. VM IP I remember, but not the old one sorry. Unfortunatley, I have created a max instances spending all my quota, I can tear it down and try again perhaps on weekend and pass the IP
I'm getting the same error on P9 machine...
The log:
[0m Build Date = Dec 13 2017 13:46:58 FW Version = buildd@ release 20170724 Press "s" to enter Open Firmware.
[0m[?25hC0000C0100C0120C0140C0200C0240C0260C02E0C0300C0320C0340C0360C0370C0380C0371C0372C0373C0374C03F0C0400C0480C04C0C04D0C0500Populating /vdevice methods Populating /vdevice/vty@30000000 Populating /vdevice/nvram@71000000 C05A0Populating /pci@800000020000000 00 0800 (D) : 1af4 1000 virtio [ net ] 00 1000 (D) : 1b36 000d serial bus [ usb-xhci ] 00 1800 (D) : 1af4 1001 virtio [ block ] 00 2000 (D) : 1af4 1002 unknown-legacy-device* 00 2800 (D) : 1234 1111 qemu vga C0600C06C0C0700C0800C0880No NVRAM common partition, re-initializing... C0890C08A0C08A8Installing QEMU fb
C08B0Scanning USB XHCI: Initializing USB Keyboard C08C0C08D0No console specified using screen & keyboard C08E0C08E8C08FF
Welcome to Open FirmwareCopyright (c) 2004, 2017 IBM Corporation All rights reserved. This program and the accompanying materials are made available under the terms of the BSD License available at http://www.opensource.org/licenses/bsd-license.php
Trying to load: from: /pci@800000020000000/scsi@3 ... Successfully loaded HV-RTAS-UPDATE error: -4 Linux ppc64le
1 SMP Debian 4.
And, the system stay on this step....
Are there any updates on expected POWER9 availability ? It seems one can only choose POWER8 as the availability zone right now, no matter which of the provided images one selects ?
@martin-frbg not yet, we need to get a dependency fixed given the issues is happening with the KVM version set for the version of OpenStack we have configured. We hope to get the P9s back soon.
my account is disabled so I have no way to validate
@rpsene do you have any news about POWER9 availability? :) As Ubuntu 22.04 only runs on POWER9+ I'm curious how your roadmap currently looks like.
@gregorkistler We've already started working on the Micloud update. In about the next two months, we're supposed to have Power9 back on MInicloud.
Great news, thank you :)
when ppc9 machine is chosen (from Availabablity Zone) it can not boot. Tried ubuntu 19.04/18.04/Debian-9, Fedora-29 seems to boot ok. Debian like systems are stuck like below in console. power8 boots ok on debian-like systems