You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@ignite.apache.org by "Andrey Khitrin (Jira)" <ji...@apache.org> on 2022/11/22 07:18:00 UTC
[jira] [Updated] (IGNITE-18218) Cluster initialization should be covered in docs better
[ https://issues.apache.org/jira/browse/IGNITE-18218?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Andrey Khitrin updated IGNITE-18218:
------------------------------------
Description:
While playing with AI3 beta1 deb package, I've noticed that 'ignite3db' service doesn't join any cluster after being installed and started. That may be fine from the DB's point of view, but this behavior may not obvious for a mere user. In order to make AI3 ready to work, user needs to run 'ignite3 cluster init ...' first. Unfortunately, this point is not currently covered properly neither in [DEVNOTES.md|https://github.com/apache/ignite-3/blob/main/DEVNOTES.md] nor in [existing docs|https://ignite.apache.org/docs/3.0.0-beta/quick-start/getting-started-guide]. Command 'systemctl status ignite3db.service' says "active (running)", command 'ignite3 node status' shows "state: starting", DB log says "Components started, joining the cluster". Everything seem to work fine, but the next user's step is not obvious here. User may try to open SQL console but it will not work until cluster is initialized.
I'm afraid a significant part of newcomers may be discouraged by this behavior. They could be a step close to the final line, but think "it doesn't work with no reason", and refuse to try AI3 anymore. To avoid this, some words about "cluster init" should be added to "Getting started guide" and, probably, to DEVNOTES.
was:
While playing with AI3 beta1 deb package, I've noticed that 'ignite3db' service doesn't join any cluster after being installed and started. That may be fine from the DB's point of view, but this behavior may not obvious for a mere user. In order to make AI3 ready to work, user needs to run 'ignite3 cluster init ...' first. Unfortunately, this point is not currently covered properly neither in [DEVNOTES.md|https://github.com/apache/ignite-3/blob/main/DEVNOTES.md] nor in [existing docs|https://ignite.apache.org/docs/3.0.0-beta/quick-start/getting-started-guide]. Command 'systemctl status ignite3db.service' says "active (running)", command 'ignite3 node status' shows "state: starting", DB log says "Components started, joining the cluster". Everything seem to work fine, but the next user's step is not obvious here. User may try to open SQL console but it will not work until cluster is initialized.
I'm afraid a significant part of newcomers may be discouraged by this behavior. They could be a step close to the final line, but think "it doesn't work with no reason", and refuse to try AI3 anymore. To avoid this, some words about "cluster init" should be added to "Getting started guide" and, probably, to DEVNOTES.
> Cluster initialization should be covered in docs better
> -------------------------------------------------------
>
> Key: IGNITE-18218
> URL: https://issues.apache.org/jira/browse/IGNITE-18218
> Project: Ignite
> Issue Type: Improvement
> Components: documentation
> Affects Versions: 3.0.0-beta1
> Reporter: Andrey Khitrin
> Priority: Major
> Labels: ignite-3
>
> While playing with AI3 beta1 deb package, I've noticed that 'ignite3db' service doesn't join any cluster after being installed and started. That may be fine from the DB's point of view, but this behavior may not obvious for a mere user. In order to make AI3 ready to work, user needs to run 'ignite3 cluster init ...' first. Unfortunately, this point is not currently covered properly neither in [DEVNOTES.md|https://github.com/apache/ignite-3/blob/main/DEVNOTES.md] nor in [existing docs|https://ignite.apache.org/docs/3.0.0-beta/quick-start/getting-started-guide]. Command 'systemctl status ignite3db.service' says "active (running)", command 'ignite3 node status' shows "state: starting", DB log says "Components started, joining the cluster". Everything seem to work fine, but the next user's step is not obvious here. User may try to open SQL console but it will not work until cluster is initialized.
> I'm afraid a significant part of newcomers may be discouraged by this behavior. They could be a step close to the final line, but think "it doesn't work with no reason", and refuse to try AI3 anymore. To avoid this, some words about "cluster init" should be added to "Getting started guide" and, probably, to DEVNOTES.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)