Merge pull request #6787 from chu11/issue6589_flux_fsck_basic

Support new flux-fsck tool to check integrity of content store

Author: mergify[bot]

doc/Makefile.am

@@ -26,6 +26,7 @@ MAN1_FILES_PRIMARY = \
 	man1/flux-getattr.1 \
 	man1/flux-dmesg.1 \
 	man1/flux-dump.1 \
+	man1/flux-fsck.1 \
 	man1/flux-archive.1 \
 	man1/flux-content.1 \
 	man1/flux-config.1 \

doc/man1/flux-fsck.rst

@@ -0,0 +1,65 @@
+============
+flux-fsck(1)
+============
+
+
+SYNOPSIS
+========
+
+**flux** **fsck** [*OPTIONS*]
+
+
+DESCRIPTION
+===========
+
+.. program flux fsck
+
+The :program:`flux fsck` checks the integrity of the KVS backing
+store, starting with the most recent checkpoint (root version) written
+to the backing store.
+
+
+OPTIONS
+=======
+
+.. option:: -h, --help
+
+   Summarize available options.
+
+.. option:: -v, --verbose
+
+   List keys on stderr as they are validated.
+
+.. option:: -q, --quiet
+
+   Don't output diagnostic messages and discovered errors.
+
+
+EXIT STATUS
+===========
+
+0
+  Content store valid
+
+1
+  One or more errors were discovered
+
+
+RESOURCES
+=========
+
+.. include:: common/resources.rst
+
+
+FLUX RFC
+========
+
+:doc:`rfc:spec_10`
+
+:doc:`rfc:spec_11`
+
+
+SEE ALSO
+========
+
+:man1:`flux-content`, :man1:`flux-kvs`

doc/manpages.py

@@ -30,6 +30,7 @@
     ('man1/flux-cron', 'flux-cron', 'Cron-like utility for Flux', [author], 1),
     ('man1/flux-dmesg', 'flux-dmesg', 'access broker ring buffer', [author], 1),
     ('man1/flux-dump', 'flux-dump', 'Write KVS snapshot to portable archive', [author], 1),
+    ('man1/flux-fsck', 'flux-fsck', 'Check integrity of KVS backing store', [author], 1),
     ('man1/flux-env', 'flux-env', 'Print the flux environment or execute a command inside it', [author], 1),
     ('man1/flux-event', 'flux-event', 'Send and receive Flux events', [author], 1),
     ('man1/flux-exec', 'flux-exec', 'Execute processes across flux ranks', [author], 1),

doc/test/spell.en.pws

@@ -968,3 +968,4 @@ UWjZ
 YQ
 composable
 orchestrator
+fsck

src/cmd/Makefile.am

@@ -79,7 +79,8 @@ flux_SOURCES = \
 	builtin/archive.c \
 	builtin/pmi.c \
 	builtin/shutdown.c \
-	builtin/lptest.c
+	builtin/lptest.c \
+	builtin/fsck.c
 nodist_flux_SOURCES = \
 	builtin-cmds.c
 dist_bin_SCRIPTS = flux-python

src/cmd/builtin/dump.c

@@ -64,22 +64,20 @@ void read_error (const char *fmt, ...)
 
 static void progress_notify (flux_t *h)
 {
-    if (sd_notify_flag) {
-        flux_future_t *f;
-        char buf[64];
-
-        snprintf (buf,
-                  sizeof (buf),
-                  "flux-dump(1) has archived %d keys",
-                  keycount);
-        f = flux_rpc_pack (h,
-                           "state-machine.sd-notify",
-                           FLUX_NODEID_ANY,
-                           FLUX_RPC_NORESPONSE,
-                           "{s:s}",
-                           "status", buf);
-        flux_future_destroy (f);
-    }
+    flux_future_t *f;
+    char buf[64];
+
+    snprintf (buf,
+              sizeof (buf),
+              "flux-dump(1) has archived %d keys",
+              keycount);
+    f = flux_rpc_pack (h,
+                       "state-machine.sd-notify",
+                       FLUX_NODEID_ANY,
+                       FLUX_RPC_NORESPONSE,
+                       "{s:s}",
+                       "status", buf);
+    flux_future_destroy (f);
 }
 
 static void progress (flux_t *h, int delta_keys)