[
https://issues.apache.org/jira/browse/HIVE-14063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15347422#comment-15347422
]
Vihang Karajgaonkar commented on HIVE-14063:
--------------------------------------------
Thanks [~szehon] for your inputs.
{quote}
One usability issue is that those beeline properties are defined by just a
letter like 'u', 'n', 'p'. Hence it would be confusing in the properties file
if the key is different, maybe its better if beeline itself took more
descriptive names in addition to letters to make it consistent.
{quote}
I am planning to use descriptive arguments instead of single letter like 'url',
'user', 'password' for 'u', 'n' and 'p' respectively. That way we remain
consistent with the existing documentation and the keys are user-friendly too.
{quote}
A lot of times there is more than one HS2 in a cluster, like HA Proxy or ZK
quorum load balancing. I assume the proposal also allows for connecting to
those URL's as well, to get a random HS2? And also for kerberos as well?
{quote}
Thanks for pointing this out. I have investigated the Kerberos environment
use-case and it should work fine with the approach above. I am still testing it
though. I will update if I find it otherwise. Regarding ZK quorum or HA proxy
based load balancing, I will need to investigate more. I will let you know.
{quote}
And one question to think about is , if you override 'url' from the command
line, will beeline prompt you for new username, password, as its a new HS2 HA
group/instance.
{quote}
If the user overrides using command line like eg. !connect, the behavior should
remain same as existing. Beeline will not look for a default properties file.
Please let me know if I misunderstood your question. Thanks.
{quote}
Also, what about renaming the file to be something like 'hive.cnf', or
'hive.conf', or 'beeline.cnf', or other different than .properties? properties
sounds more like a Java properties file than a Hive configuration file.
{quote}
I think we should use beeline.cnf or beeline.conf in that case. hive.conf or
hive.cnf could be confused with HiveCli. Its better to have beeline in the file
name so that users will not get confused about the purpose of the file.
[~spena] Let me know what do you think?
> beeline to auto connect to the HiveServer2
> ------------------------------------------
>
> Key: HIVE-14063
> URL: https://issues.apache.org/jira/browse/HIVE-14063
> Project: Hive
> Issue Type: Improvement
> Components: Beeline
> Reporter: Vihang Karajgaonkar
> Assignee: Vihang Karajgaonkar
> Priority: Minor
>
> Currently one has to give an jdbc:hive2 url in order for Beeline to connect a
> hiveserver2 instance. It would be great if Beeline can get the info somehow
> (from a properties file at a well-known location?) and connect automatically
> if user doesn't specify such a url. If the properties file is not present,
> then beeline would expect user to provide the url and credentials using
> !connect or ./beeline -u .. commands
> While Beeline is flexible (being a mere JDBC client), most environments would
> have just a single HS2. Having users to manually connect into this via either
> "beeline ~/.propsfile" or -u or !connect statements is lowering the
> experience part.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
