Commits on Apr 3, 2019

1. cli: actually fetch per-node tables from nodes …

   The previous code used the sql connection to the first node to fetch
   all data. Unfortunately, we have to guess the SQL connection string
   for the other nodes (and may not get it right), but I've verified
   that this doesn't fail out the whole `debug zip` process. In the
   process of verifying that (running with one node down) I also fixed
   a few buglets related to that, including creating files multiple
   times and generally confused-looking output.
   
   Release note: None

   

   tbg committed Apr 3, 2019
   Configuration menu

   • View commit details
   • Copy full SHA for cf3ca4f
   • Browse repository at this point

   Copy the full SHA

   cf3ca4f View commit details

   Browse the repository at this point in the history

2. cli: enforce inclusion of new crdb_internal tables in debug zip …

   This is achieved via a unit test. I went through all of the missing
   tables and made judgement calls on whether they needed to be included.
   
   This resulted in the inclusion of some obvious earlier omissions.
   
   I also went through the crdb_internal tables and added a naming
   convention and tried to add TODOs to all methods violating it.
   It's a bad idea to rename any of these; we should probably alias
   them to their desired name and leave the old name valid, somehow
   without listing it in `SHOW TABLES FROM crdb_internal`.
   
   Release note (bug fix): Data that was previously omitted from `debug
   zip` is now included.

   

   tbg committed Apr 3, 2019
   Configuration menu

   • View commit details
   • Copy full SHA for 81082f7
   • Browse repository at this point

   Copy the full SHA

   81082f7 View commit details

   Browse the repository at this point in the history