Flatten the 'db' commands, e.g. 'db prefix' to 'db/prefix'

The user experience with commands that have sub-commands is not great
and it makes it harder to implement completion in 'cilium-dbg shell'
if we so choose. There isn't really any good rational for nesting
the commands like this, so let's go with Plan 9 style command names
(https://9p.io/magic/man2html/8/auth), e.g. "db prefix" becomes
"db/prefix".

Signed-off-by: Jussi Maki <[email protected]>

Author: joamaki

reconciler/testdata/batching.txtar

@@ -7,38 +7,38 @@ start-reconciler with-batchops
 # From here this is the same as incremental.txtar.
 
 # Step 1: Insert non-faulty objects
-db insert test-objects obj1.yaml
-db insert test-objects obj2.yaml
-db insert test-objects obj3.yaml
-db cmp test-objects step1+3.table
+db/insert test-objects obj1.yaml
+db/insert test-objects obj2.yaml
+db/insert test-objects obj3.yaml
+db/cmp test-objects step1+3.table
 expect-ops update(1) update(2) update(3)
 
 # Reconciler should be running and reporting health
 health 'job-reconcile.*level=OK.*message=OK, 3 object'
 
 # Step 2: Update object '1' to be faulty and check that it fails and is being
 # retried.
-db insert test-objects obj1_faulty.yaml
+db/insert test-objects obj1_faulty.yaml
 expect-ops 'update(1) fail' 'update(1) fail' 
-db cmp test-objects step2.table
+db/cmp test-objects step2.table
 health 'job-reconcile.*level=Degraded.*1 error'
 
 # Step 3: Set object '1' back to healthy state
-db insert test-objects obj1.yaml
+db/insert test-objects obj1.yaml
 expect-ops 'update(1)'
-db cmp test-objects step1+3.table
+db/cmp test-objects step1+3.table
 health 'job-reconcile.*level=OK'
 
 # Step 4: Delete '1' and '2'
-db delete test-objects obj1.yaml
-db delete test-objects obj2.yaml
-db cmp test-objects step4.table
+db/delete test-objects obj1.yaml
+db/delete test-objects obj2.yaml
+db/cmp test-objects step4.table
 expect-ops 'delete(1)' 'delete(2)'
 
 # Step 5: Try to delete '3' with faulty target
 set-faulty true
-db delete test-objects obj3.yaml
-db cmp test-objects empty.table
+db/delete test-objects obj3.yaml
+db/cmp test-objects empty.table
 expect-ops 'delete(3) fail'
 health 'job-reconcile.*level=Degraded.*1 error'
 

reconciler/testdata/incremental.txtar

@@ -5,39 +5,39 @@ hive start
 start-reconciler
 
 # Step 1: Insert non-faulty objects
-db insert test-objects obj1.yaml
-db insert test-objects obj2.yaml
-db insert test-objects obj3.yaml
-db cmp test-objects step1+3.table
+db/insert test-objects obj1.yaml
+db/insert test-objects obj2.yaml
+db/insert test-objects obj3.yaml
+db/cmp test-objects step1+3.table
 expect-ops update(1) update(2) update(3)
 
 # Reconciler should be running and reporting health
 health 'job-reconcile.*level=OK.*message=OK, 3 object'
 
 # Step 2: Update object '1' to be faulty and check that it fails and is being
 # retried.
-db insert test-objects obj1_faulty.yaml
-db cmp test-objects step2.table
+db/insert test-objects obj1_faulty.yaml
+db/cmp test-objects step2.table
 expect-ops 'update(1) fail' 'update(1) fail' 
 health 'job-reconcile.*level=Degraded.*1 error'
 
 # Step 3: Set object '1' back to healthy state
-db insert test-objects obj1.yaml
-db show test-objects
-db cmp test-objects step1+3.table
+db/insert test-objects obj1.yaml
+db/show test-objects
+db/cmp test-objects step1+3.table
 expect-ops 'update(1)'
 health 'job-reconcile.*level=OK'
 
 # Step 4: Delete '1' and '2'
-db delete test-objects obj1.yaml
-db delete test-objects obj2.yaml
-db cmp test-objects step4.table
+db/delete test-objects obj1.yaml
+db/delete test-objects obj2.yaml
+db/cmp test-objects step4.table
 expect-ops 'delete(1)' 'delete(2)'
 
 # Step 5: Try to delete '3' with faulty target
 set-faulty true
-db delete test-objects obj3.yaml
-db cmp test-objects empty.table
+db/delete test-objects obj3.yaml
+db/cmp test-objects empty.table
 expect-ops 'delete(3) fail'
 health 'job-reconcile.*level=Degraded.*1 error'
 

reconciler/testdata/pruning.txtar

@@ -2,10 +2,10 @@ hive start
 start-reconciler with-prune
 
 # Pruning without table being initialized does nothing.
-db insert test-objects obj1.yaml
+db/insert test-objects obj1.yaml
 expect-ops update(1)
 prune
-db insert test-objects obj2.yaml 
+db/insert test-objects obj2.yaml 
 expect-ops update(2) update(1)
 health 'job-reconcile.*level=OK'
 
@@ -33,12 +33,12 @@ expvar
 grep 'prune_current_errors.test: 0'
 
 # Delete an object and check pruning happens without it
-db delete test-objects obj1.yaml
+db/delete test-objects obj1.yaml
 prune
 expect-ops 'prune(n=1)' delete(1)
 
 # Prune without objects
-db delete test-objects obj2.yaml
+db/delete test-objects obj2.yaml
 prune
 expect-ops prune(n=0) delete(2) prune(n=1)
 

reconciler/testdata/refresh.txtar

@@ -2,27 +2,27 @@ hive start
 start-reconciler with-refresh
 
 # Step 1: Add a test object.
-db insert test-objects obj1.yaml
+db/insert test-objects obj1.yaml
 expect-ops 'update(1)'
-db cmp test-objects step1.table
+db/cmp test-objects step1.table
 
 # Step 2: Set the object as updated in the past to force refresh
-db insert test-objects obj1_old.yaml
+db/insert test-objects obj1_old.yaml
 expect-ops 'update-refresh(1)'
 
 # Step 3: Refresh with faulty target, should see fail & retries
 set-faulty true
-db insert test-objects obj1_old.yaml
+db/insert test-objects obj1_old.yaml
 expect-ops 'update-refresh(1) fail' 'update-refresh(1) fail'
-db cmp test-objects step3.table
+db/cmp test-objects step3.table
 health
 health 'job-reconcile.*Degraded'
 
 # Step 4: Back to health
 set-faulty false
-db insert test-objects obj1_old.yaml
+db/insert test-objects obj1_old.yaml
 expect-ops 'update-refresh(1)'
-db cmp test-objects step4.table
+db/cmp test-objects step4.table
 health 'job-reconcile.*OK, 1 object'
 
 # -----