Managix Documentation: Multi-node set up needs two fixes

[GoogleCodeExporter]

The installation guide for a multi-node AsterixDB setup needs the following 
fixes.
a)  change of attribute name: log_Dir --> log_dir
b) inclusion and description of the attribute txn_log_dir

Original issue reported on code.google.com by RamanGro...@gmail.com on 4 Feb 2014 at 3:56

[GoogleCodeExporter]

Original comment by ima...@uci.edu on 14 Oct 2014 at 9:34

• Added labels: Weekly

[GoogleCodeExporter]

I'll adopt this.

Original comment by c...@lambda.nu on 23 Oct 2014 at 12:43

[GoogleCodeExporter]

FYI, these two particular issues seem to be resolved in the documentation (and 
have been since June 2013, oddly). I am going through the entire tutorial and 
have found a few other similar errors (cluster-ip instead of cluster_ip, etc). 
I will propose a change with all my findings tonight.

Original comment by c...@lambda.nu on 24 Oct 2014 at 12:39

[GoogleCodeExporter]

I missed that the initial error report specifically called out the multi-node 
setup documentation. Indeed both errors existed there (although the erroneous 
element was called "logdir" rather than "log_Dir").

At any rate, http://fulliautomatix.ics.uci.edu:8443/#/c/154/2 fixes these 
critical errors and several trivial errors in the installer doc. I also 
reorganized the doc slightly to flow better - previously, in the multi-node 
configuration, the section on configuring ssh came after creating and 
validating the cluster configuration file, but "managix validate -c 
cluster.xml" depends on password-less ssh already being set up. I moved the ssh 
discussion first and added a little more detail.

I also added a check for the existence of <txn_log_dir> to "managix validate 
-c".

Original comment by c...@lambda.nu on 24 Oct 2014 at 8:34

[GoogleCodeExporter]

Original comment by c...@lambda.nu on 24 Oct 2014 at 5:11

• Changed state: Verified