Summary table column spacing is not optimal

wavectl show alert コマンドから始め、--name Kubernetes の有無で出力を比較します。サーバーの結果がクライアント側で絞り込まれる箇所と、サマリーテーブルの間隔が計算される箇所を追跡します。表示される列が除外されたアラートではなく最終的にフィルタリングされたアラート集合を使用し、表示されるテーブル出力を維持できれば完了です。

If the user has used resource options other than the --customerTag, the column spacing can be suboptimal.

For example see here

$ ./wavectl show alert --name Kubernetes
ID               NAME                                                                                     STATUS                              SEVERITY
1523082347619    Kubernetes - Node Network Utilization - HIGH (Prod)                                      CHECKING                            WARN
1523082347824    Kubernetes - Node Cpu Utilization - HIGH (Prod)                                          CHECKING                            WARN
1523082348005    Kubernetes - Node Memory Swap Utilization - HIGH (Prod)                                  SNOOZED                             WARN

There is so much space between the NAME column and STATUS columns. The reason is that the column spacing is calculated using the alerts without considering the effects of other locally executed filters. In this example the --name Kubernetes command line param.

Seeing the full set of alerts, (without the --name Kubernetes option) shows the reason of the separation

hbaba@mbp-005063:~/box/src/skynet/wavectl/doc/bin (master)$ ./wavectl show alert
ID               NAME                                                                                     STATUS                              SEVERITY
1523082347619    Kubernetes - Node Network Utilization - HIGH (Prod)                                      CHECKING                            WARN
1523082347824    Kubernetes - Node Cpu Utilization - HIGH (Prod)                                          CHECKING                            WARN
1523082348005    Kubernetes - Node Memory Swap Utilization - HIGH (Prod)                                  SNOOZED                             WARN
1523082348172    Wavefront Freshness                                                                      CHECKING                            WARN
1523082348348    Legacy admin console 500sTODO_SOME_TEST_12432345_STRINGTODO_SOME_TEST_12432345_STRING    CHECKING                            WARN
1523082348512    MySQL Live - Running Kernel Matches Boot Kernel - FALSE                                  IN_MAINTENANCE CHECKING             WARN
1523082348708    Orion Response time more than 2 seconds                                                  FIRING                              INFO
1523082348877    Realtime - Aggregate FetchMessage Median Latency - HIGH                                  CHECKING                            SEVERE
1523082349048    Updates Page - Total ReverseFetchMessage Median Latency - HIGH                           CHECKING                            SEVERE
1523082349224    CacheManager Exception Rate - STAGING                                                    CHECKING                            SEVERE
1523082349397    Prod - Office Online Unexpected Errors - HIGH                                            CHECKING                            SEVERE

There is an alert that has a really long name and even though that is not in the final displayed list, it caused the columns to be too widely separated.

All alerts matching the --customerTag are received from the Wavefront server, and they are further narrowed down on the wavectl client . The show summary table column separation should use the final set of alerts.