Closed alexellis closed 4 years ago
Provided sample vars works fine for default Ubuntu 18.04 packet installation. The issue is in the script itself:
https://github.com/tinkerbell/tink/pull/62#issuecomment-617229081
@gauravgahlot already updated the script to use ifconfig and I just tested it so I can confirm that supplied example vars works fine. Feel free to test yourself and let me know if it works for you too
Thanks for helping with this. Can you elaborate on what you mean by "it works fine"? When I ran through the steps and it did not? Do you mean it will work fine when #62 is merged (or now that it's been updated)?
By "it works fine" I mean you no longer hit "unknown interface" error and script will fail on a later step (logging to docker registry) due to different error. And I mean when #62 gets merged of course. But I'm seeing few more issues with networking which I guess makes docker login fail, so while the example configuration is fine - it won't work until all the issues in the script get resolved.
Thank you for clarifying, so I am not going crazy 😄
@alexellis I tested the script for CentOS 7 and Ubuntu 18.04 on:
c3.small.x86
export TB_INTERFACE=enp1s0f1
export TB_NETWORK=192.168.1.0/29
export TB_IPADDR=192.168.1.1
export TB_REGUSER=admin
Executed the script with:
$ curl https://raw.githubusercontent.com/infracloudio/tink/deploy_stack/setup.sh | bash
Added the same to PR description and docs as well: https://github.com/tinkerbell/tink/pull/62/files#diff-85ad37f4d4d62ce0f24ea8a26b95ba09R65
I'm trying it now.
I can confirm that the script completed, but there is't a clear next action.
Creating deploy_kibana_1 ... done
WARNING: The PACKET_API_AUTH_TOKEN variable is not set. Defaulting to a blank string.
WARNING: The PACKET_API_URL variable is not set. Defaulting to a blank string.
Name Command State Ports
-------------------------------------------------------------------------------------------------------------------------------------
deploy_boots_1 /boots -dhcp-addr 0.0.0.0: ... Up
deploy_cacher_1 /cacher Up 0.0.0.0:42111->42111/tcp, 0.0.0.0:42112->42112/tcp
deploy_certs_1 /entrypoint.sh Exit 0
deploy_db_1 docker-entrypoint.sh postgres Up (healthy) 0.0.0.0:5432->5432/tcp
deploy_elasticsearch_1 /usr/local/bin/docker-entr ... Up 0.0.0.0:9200->9200/tcp, 0.0.0.0:9300->9300/tcp
deploy_fluentbit_1 /fluent-bit/bin/fluent-bit ... Up 2020/tcp, 0.0.0.0:24224->24224/tcp, 0.0.0.0:24224->24224/udp
deploy_hegel_1 cmd/hegel Up
deploy_kibana_1 /usr/local/bin/kibana-docker Up 0.0.0.0:5601->5601/tcp
deploy_nginx_1 nginx -g daemon off; Up 192.168.1.2:80->80/tcp
deploy_registry_1 /entrypoint.sh /etc/docker ... Up
deploy_tink-cli_1 /bin/sh -c sleep infinity Up
deploy_tink-server_1 tink-server Up (healthy) 0.0.0.0:42113->42113/tcp, 0.0.0.0:42114->42114/tcp
INFO: tinkerbell stack setup completed successfully on ubuntu server
root@ams1-c3:~#
I see that the tink CLI's source code is downloaded under tink/cli/tink, but it's not built? Do you plan on downloading a release binary or building as part of the script so that users can make use of it after bootstrap?
A help message with what to do next would also be a good addition at the end of the script.
This error is also confusing / unactionable. What is the value when used with compose? It's not documented here -> https://github.com/tinkerbell/tink/blob/master/docs/first-good-workflow.md or here -> https://github.com/tinkerbell/tink/blob/master/docs/setup.md
root@ams1-c3:~/tink/cli/tink# ./tink
2020/04/23 10:01:51 undefined TINKERBELL_CERT_URL
I think you still need to use tink from tink-cli
container. I was using alias for it alias tink='PACKET_API_AUTH_TOKEN="" PACKET_API_URL="" docker-compose -f /root/deploy/docker-compose.yml exec tink-cli tink'
As soon as the PR #62 is merged, I will be sending another one with changes to first-good-workflow. And accordingly will also add a "next" section at the end of setup script.
@alexellis Could you comment on the remaining delta between this request and what you now see at https://tinkerbell.org/setup please?
We believe @kdeng3849 and others may have closed the gap but want to be sure we haven't missed anything out.
@gauravgahlot #62 is merged now, did you have time to raise the PR you mentioned for the "next action" section at the end of the script?
@alexellis I had updated the first-good-workflow
docs (PR #85). I have sent another PR #87 for updating the setup script and docs with "what's next".
@mrmrcoleman WRT https://tinkerbell.org and the issue, I have done work. Since we could not take that up in the call today, I will be sending a few details on the channel.
Thanks @gauravgahlot I would check @jacobsmith928 's comment in Slack about that issue, as his team are already working on updating the website.
Additional feedback given on https://github.com/tinkerbell/tinkerbell.github.io/issues/6
@alexellis Since we have other issues that describe current challenges with the setup, can we close this one now?
This is covered now. At the moment setup.sh and the terraform both work and cover this.
I would like to see a known configuration for the installer script in #62 which provides everything needed without any guess-work or troubleshooting.
I.e. "run these commands" and it results in Tinkerbell being up.
cc @gauravgahlot
A comment on this issue should be sufficient to get started, but the above should be in the docs eventually.