Implement "deleted" flag

docs/usage/cli.en.md

@@ -146,6 +146,7 @@ Rename, resize image or change its readonly status. Images with children can't b
 If the new size is smaller than the old size, extra data will be purged.
 You should resize file system in the image, if present, before shrinking it.
 
+* `--deleted 1|0` - Set/clear 'deleted image' flag (set automatically during unfinished deletes).
 * `-f|--force` - Proceed with shrinking or setting readwrite flag even if the image has children.
 * `--down-ok` - Proceed with shrinking even if some data will be left on unavailable OSDs.
 

docs/usage/cli.ru.md

@@ -149,6 +149,7 @@ vitastor-cli snap-create [-p|--pool <id|name>] <image>@<snapshot>
 Если новый размер меньше старого, "лишние" данные будут удалены, поэтому перед уменьшением
 образа сначала уменьшите файловую систему в нём.
 
+* `--deleted 1|0` - Установить/снять флаг "образ удалён" (устанавливается при незавершённом удалении).
 * `-f|--force` - Разрешить уменьшение или перевод в чтение-запись образа, у которого есть клоны.
 * `--down-ok` - Разрешить уменьшение, даже если часть данных останется неудалённой на недоступных OSD.
 

mon/etcd_schema.js

@@ -216,6 +216,7 @@ const etcd_tree = {
                     parent_pool?: <pool_id>,
                     parent_id?: <inode_t>,
                     readonly?: boolean,
+                    deleted?: boolean,
                 }
             }
         }, */

src/client/etcd_state_client.cpp

@@ -1217,6 +1217,7 @@ void etcd_state_client_t::parse_state(const etcd_kv_t & kv)
                     .size = value["size"].uint64_value(),
                     .parent_id = parent_inode_num,
                     .readonly = value["readonly"].bool_value(),
+                    .deleted = value["deleted"].bool_value(),
                     .meta = value["meta"],
                     .mod_revision = kv.mod_revision,
                 });
@@ -1305,6 +1306,10 @@ json11::Json::object etcd_state_client_t::serialize_inode_cfg(inode_config_t *cf
     {
         new_cfg["readonly"] = true;
     }
+    if (cfg->deleted)
+    {
+        new_cfg["deleted"] = true;
+    }
     if (cfg->meta.is_object())
     {
         new_cfg["meta"] = cfg->meta;

src/client/etcd_state_client.h

@@ -72,6 +72,7 @@ struct inode_config_t
     uint64_t size = 0;
     inode_t parent_id = 0;
     bool readonly = false;
+    bool deleted = false;
     // Arbitrary metadata
     json11::Json meta;
     // Change revision of the metadata in etcd

src/cmd/cli.cpp

@@ -51,8 +51,9 @@ static const char* help_text =
     "  Rename, resize image or change its readonly status. Images with children can't be made read-write.\n"
     "  If the new size is smaller than the old size, extra data will be purged.\n"
     "  You should resize file system in the image, if present, before shrinking it.\n"
-    "  -f|--force  Proceed with shrinking or setting readwrite flag even if the image has children.\n"
-    "  --down-ok   Proceed with shrinking even if some data will be left on unavailable OSDs.\n"
+    "  --deleted 1|0  Set/clear 'deleted image' flag (set automatically during unfinished deletes).\n"
+    "  -f|--force     Proceed with shrinking or setting readwrite flag even if the image has children.\n"
+    "  --down-ok      Proceed with shrinking even if some data will be left on unavailable OSDs.\n"
     "\n"
     "vitastor-cli dd [iimg=<image> | if=<file>] [oimg=<image> | of=<file>] [bs=1M]\n"
     "    [count=N] [seek/oseek=N] [skip/iseek=M] [iodepth=N] [status=progress]\n"

src/cmd/cli_ls.cpp

@@ -371,7 +371,8 @@ resume_1:
                 kv.second["delete_q"] = format_q(kv.second["delete_queue"].number_value());
             }
             kv.second["size_fmt"] = format_size(kv.second["size"].uint64_value());
-            kv.second["ro"] = kv.second["readonly"].bool_value() ? "RO" : "-";
+            kv.second["ro"] = kv.second["deleted"].bool_value() ? "DEL" :
+                (kv.second["readonly"].bool_value() ? "RO" : "-");
         }
         result.text = print_table(to_list(), cols, parent->color);
         state = 100;

src/cmd/cli_modify.cpp

@@ -4,6 +4,7 @@
 #include "cli.h"
 #include "cluster_client.h"
 #include "str_util.h"
+#include "json_util.h"
 
 // Rename, resize image (and purge extra data on shrink) or change its readonly status
 struct image_changer_t
@@ -15,6 +16,7 @@ struct image_changer_t
     uint64_t new_size = 0;
     bool force_size = false, inc_size = false;
     bool set_readonly = false, set_readwrite = false, force = false;
+    bool set_deleted = false, new_deleted = false;
     bool down_ok = false;
     // interval between fsyncs
     int fsync_interval = 128;
@@ -82,6 +84,7 @@ struct image_changer_t
         }
         if ((!set_readwrite || !cfg.readonly) &&
             (!set_readonly || cfg.readonly) &&
+            (!set_deleted || cfg.deleted == new_deleted) &&
             (!new_size && !force_size || cfg.size == new_size || cfg.size >= new_size && inc_size) &&
             (new_name == "" || new_name == image_name))
         {
@@ -141,6 +144,10 @@ resume_1:
                 return;
             }
         }
+        if (set_deleted)
+        {
+            cfg.deleted = new_deleted;
+        }
         if (new_name != "")
         {
             cfg.name = new_name;
@@ -251,6 +258,8 @@ std::function<bool(cli_result_t &)> cli_tool_t::start_modify(json11::Json cfg)
     changer->force = cfg["force"].bool_value();
     changer->set_readonly = cfg["readonly"].bool_value();
     changer->set_readwrite = cfg["readwrite"].bool_value();
+    changer->set_deleted = !cfg["deleted"].is_null();
+    changer->new_deleted = json_is_true(cfg["deleted"]);
     changer->fsync_interval = cfg["fsync_interval"].uint64_value();
     if (!changer->fsync_interval)
         changer->fsync_interval = 128;

src/cmd/cli_rm.cpp

@@ -95,6 +95,10 @@ struct snap_remover_t
             goto resume_7;
         else if (state == 8)
             goto resume_8;
+        else if (state == 9)
+            goto resume_9;
+        else if (state == 10)
+            goto resume_10;
         else if (state == 100)
             goto resume_100;
         assert(!state);
@@ -129,23 +133,9 @@ resume_1:
                 continue;
             rebased_images.push_back(parent->cli->st_cli.inode_config.at(merge_children[current_child]).name);
             start_merge_child(merge_children[current_child], merge_children[current_child]);
-            if (state == 100)
-                return;
 resume_2:
-            while (!cb(result))
-            {
-                state = 2;
+            while (!wait_result(2))
                 return;
-            }
-            cb = NULL;
-            if (result.err)
-            {
-                result.data = my_result(result.data);
-                state = 100;
-                return;
-            }
-            else if (parent->progress)
-                printf("%s\n", result.text.c_str());
             parent->change_parent(merge_children[current_child], new_parent, &result);
             state = 3;
 resume_3:
@@ -164,42 +154,19 @@ resume_3:
         if (inverse_child != 0)
         {
             start_merge_child(inverse_child, inverse_parent);
-            if (state == 100)
-                return;
 resume_4:
-            while (!cb(result))
-            {
-                state = 4;
+            while (!wait_result(4))
                 return;
-            }
-            cb = NULL;
-            if (result.err)
-            {
-                result.data = my_result(result.data);
-                state = 100;
+            // Mark child as deleted
+            start_mark_deleted(inverse_child);
+resume_9:
+            while (!wait_result(9))
                 return;
-            }
-            else if (parent->progress)
-                printf("%s\n", result.text.c_str());
             // Delete "inverse" child data
             start_delete_source(inverse_child);
-            if (state == 100)
-                return;
 resume_5:
-            while (!cb(result))
-            {
-                state = 5;
+            while (!wait_result(5))
                 return;
-            }
-            cb = NULL;
-            if (result.err)
-            {
-                result.data = my_result(result.data);
-                state = 100;
-                return;
-            }
-            else if (parent->progress)
-                printf("%s\n", result.text.c_str());
             // Delete "inverse" child metadata, rename parent over it,
             // and also change parent links of the previous "inverse" child
             rename_inverse_parent();
@@ -221,22 +188,15 @@ resume_6:
                     deleted_images.push_back(parent_it->second.name);
                 deleted_ids.push_back(chain_list[current_child]);
             }
+            // Mark child as deleted
+            start_mark_deleted(chain_list[current_child]);
+resume_10:
+            while (!wait_result(10))
+                return;
             start_delete_source(chain_list[current_child]);
 resume_7:
-            while (!cb(result))
-            {
-                state = 7;
+            while (!wait_result(7))
                 return;
-            }
-            cb = NULL;
-            if (result.err)
-            {
-                result.data = my_result(result.data);
-                state = 100;
-                return;
-            }
-            else if (parent->progress)
-                printf("%s\n", result.text.c_str());
             delete_inode_config(chain_list[current_child]);
             if (state == 100)
                 return;
@@ -256,6 +216,26 @@ resume_100:
         return;
     }
 
+    bool wait_result(int base_state)
+    {
+        if (state == 100 || !cb)
+            return true;
+        while (!cb(result))
+        {
+            state = base_state;
+            return false;
+        }
+        cb = NULL;
+        if (result.err)
+        {
+            result.data = my_result(result.data);
+            state = 100;
+        }
+        else if (parent->progress)
+            printf("%s\n", result.text.c_str());
+        return true;
+    }
+
     json11::Json my_result(json11::Json src)
     {
         auto obj = src.object_items();
@@ -667,6 +647,27 @@ resume_100:
         });
     }
 
+    void start_mark_deleted(inode_t inode)
+    {
+        auto ino_it = parent->cli->st_cli.inode_config.find(inode);
+        if (ino_it == parent->cli->st_cli.inode_config.end())
+        {
+            char buf[1024];
+            snprintf(buf, 1024, "Inode 0x%jx disappeared", inode);
+            result = (cli_result_t){ .err = EIO, .text = std::string(buf) };
+            state = 100;
+            return;
+        }
+        if (ino_it->second.deleted)
+        {
+            return;
+        }
+        cb = parent->start_modify(json11::Json::object {
+            { "image", ino_it->second.name },
+            { "deleted", true },
+        });
+    }
+
     void start_delete_source(inode_t inode)
     {
         auto source = parent->cli->st_cli.inode_config.find(inode);