Redirect startup.log to stdout

[roypaulin]

In the v2 vertica image, we can now redirect startup.log to stdout by setting VERTICA_STARTUP_LOG_DUPLICATE to /proc/1/fd/1.

An example of a Vertica pod log output via kubectl logs v-sc0-0:

2023/11/06 14:34:09 New NodeManagementAgent starting
2023/11/06 14:34:09 Checking for existence of directory  /opt/vertica/log
2023/11/06 14:34:09 Moving working directory to  /opt/vertica/log
2023/11/06 14:34:09 Successfully opened file node_management_agent.log. Setting log output to that file.
{
  "goal" : 27,
  "node" : "v_vertdb_node0001",
  "progress" : 0,
  "stage" : "Reading Catalog",
  "text" : "Indexing Objects",
  "timestamp" : "2023-11-06 14:34:18.414"
}
{
  "goal" : 27,
  "node" : "v_vertdb_node0001",
  "progress" : 27,
  "stage" : "Reading Catalog",
  "text" : "Indexing Objects",
  "timestamp" : "2023-11-06 14:34:18.414"
}
{
  "node" : "v_vertdb_node0001",
  "stage" : "Connecting to Spread",
  "text" : "Connecting to spread /tmp/4803",
  "timestamp" : "2023-11-06 14:34:18.437"
}
........
{
  "node" : "v_vertdb_node0001",
  "stage" : "Waiting for Cluster Invite",
  "text" : "Prepare to be invited",
  "timestamp" : "2023-11-06 14:34:20.003"
}
{
  "node" : "v_vertdb_node0001",
  "stage" : "Waiting for Cluster Invite",
  "text" : "Ready to be invited",
  "timestamp" : "2023-11-06 14:34:20.004"
}
{
  "goal" : 1,
  "node" : "v_vertdb_node0001",
  "progress" : 1,
  "stage" : "Invite Cluster",
  "text" : "All nodes present",
  "timestamp" : "2023-11-06 14:34:20.007"
}
{
  "node" : "v_vertdb_node0001",
  "stage" : "Invite Cluster",
  "text" : "Inviting cluster",
  "timestamp" : "2023-11-06 14:34:20.009"
}
{
  "node" : "v_vertdb_node0001",
  "stage" : "Waiting for Cluster Invite",
  "text" : "Invited",
  "timestamp" : "2023-11-06 14:34:23.959"
}
{
  "node" : "v_vertdb_node0001",
  "stage" : "Startup Complete",
  "text" : "Node is UP",
  "timestamp" : "2023-11-06 14:34:24.143"
}

[spilchen]

This looks good. I think we should add changie entry for this change.