-/* Recursively descend the filesystem and visit each file. */
+/* Cumulus: Smart Filesystem Backup to Dumb Servers
+ *
+ * Copyright (C) 2006-2008 The Regents of the University of California
+ * Written by Michael Vrable <mvrable@cs.ucsd.edu>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License along
+ * with this program; if not, write to the Free Software Foundation, Inc.,
+ * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
+ */
+
+/* Main entry point for LBS. Contains logic for traversing the filesystem and
+ * constructing a backup. */
#include <dirent.h>
#include <errno.h>
#include "remote.h"
#include "store.h"
#include "sha1.h"
+#include "subfile.h"
#include "util.h"
using std::list;
/* Look up this file in the old stat cache, if we can. If the stat
* information indicates that the file has not changed, do not bother
- * re-reading the entire contents. */
+ * re-reading the entire contents. Even if the information has been
+ * changed, we can use the list of old blocks in the search for a sub-block
+ * incremental representation. */
bool cached = false;
+ list<ObjectReference> old_blocks;
- if (metawriter->find(path) && metawriter->is_unchanged(&stat_buf)) {
+ bool found = metawriter->find(path);
+ if (found)
+ old_blocks = metawriter->get_blocks();
+
+ if (found && metawriter->is_unchanged(&stat_buf)) {
cached = true;
- list<ObjectReference> blocks = metawriter->get_blocks();
/* If any of the blocks in the object have been expired, then we should
* fall back to fully reading in the file. */
- for (list<ObjectReference>::const_iterator i = blocks.begin();
- i != blocks.end(); ++i) {
+ for (list<ObjectReference>::const_iterator i = old_blocks.begin();
+ i != old_blocks.end(); ++i) {
const ObjectReference &ref = *i;
if (!db->IsAvailable(ref)) {
cached = false;
/* If everything looks okay, use the cached information */
if (cached) {
file_info["checksum"] = metawriter->get_checksum();
- for (list<ObjectReference>::const_iterator i = blocks.begin();
- i != blocks.end(); ++i) {
+ for (list<ObjectReference>::const_iterator i = old_blocks.begin();
+ i != old_blocks.end(); ++i) {
const ObjectReference &ref = *i;
object_list.push_back(ref.to_string());
if (ref.is_normal())
* time. */
if (!cached) {
SHA1Checksum hash;
+ Subfile subfile(db);
+ subfile.load_old_blocks(old_blocks);
+
while (true) {
ssize_t bytes = file_read(fd, block_buf, LBS_BLOCK_SIZE);
if (bytes == 0)
ref = db->FindObject(block_csum, bytes);
}
+ list<ObjectReference> refs;
+
// Store a copy of the object if one does not yet exist
if (ref.is_null()) {
LbsObject *o = new LbsObject;
status = "new";
}
- o->set_data(block_buf, bytes);
- o->write(tss);
- ref = o->get_ref();
- db->StoreObject(ref, block_csum, bytes, block_age);
- ref.set_range(0, bytes);
- delete o;
+ subfile.analyze_new_block(block_buf, bytes);
+ refs = subfile.create_incremental(tss, o, block_age);
+ } else {
+ refs.push_back(ref);
}
- object_list.push_back(ref.to_string());
- if (ref.is_normal())
- add_segment(ref.get_segment());
- db->UseObject(ref);
+ while (!refs.empty()) {
+ ref = refs.front(); refs.pop_front();
+ object_list.push_back(ref.to_string());
+ if (ref.is_normal())
+ add_segment(ref.get_segment());
+ db->UseObject(ref);
+ }
size += bytes;
if (status == NULL)
for (list<string>::iterator i = includes.begin();
i != includes.end(); ++i) {
if (path == *i) {
- printf("Including %s\n", path.c_str());
include = true;
}
}
for (list<string>::iterator i = excludes.begin();
i != excludes.end(); ++i) {
if (path == *i) {
- printf("Excluding %s\n", path.c_str());
include = false;
}
}
for (list<string>::iterator i = searches.begin();
i != searches.end(); ++i) {
if (path == *i) {
- printf("Scanning %s\n", path.c_str());
scan_only = true;
}
}
* themselves are excluded from being backed up. */
void add_include(const char *path)
{
- printf("Add: %s\n", path);
/* Was an absolute path specified? If so, we'll need to start scanning
* from the root directory. Make sure that the user was consistent in
* providing either all relative paths or all absolute paths. */
return 1;
}
- // Dump paths for debugging/informational purposes
- {
- list<string>::const_iterator i;
-
- printf("LBS Version: %s\n", lbs_version);
-
- printf("--dest=%s\n--localdb=%s\n--upload-script=%s\n",
- backup_dest.c_str(), localdb_dir.c_str(), backup_script.c_str());
-
- printf("Includes:\n");
- for (i = includes.begin(); i != includes.end(); ++i)
- printf(" %s\n", i->c_str());
-
- printf("Excludes:\n");
- for (i = excludes.begin(); i != excludes.end(); ++i)
- printf(" %s\n", i->c_str());
-
- printf("Searching:\n");
- for (i = searches.begin(); i != searches.end(); ++i)
- printf(" %s\n", i->c_str());
- }
-
block_buf = new char[LBS_BLOCK_SIZE];
/* Initialize the remote storage layer. If using an upload script, create
if (backup_scheme.size() > 0)
checksum_filename += backup_scheme + "-";
checksum_filename = checksum_filename + desc_buf + "." + csum_type + "sums";
- RemoteFile *checksum_file = remote->alloc_file(checksum_filename);
+ RemoteFile *checksum_file = remote->alloc_file(checksum_filename,
+ "checksums");
FILE *checksums = fdopen(checksum_file->get_fd(), "w");
for (std::set<string>::iterator i = segment_list.begin();
desc_filename += backup_scheme + "-";
desc_filename = desc_filename + desc_buf + ".lbs";
- RemoteFile *descriptor_file = remote->alloc_file(desc_filename);
+ RemoteFile *descriptor_file = remote->alloc_file(desc_filename,
+ "snapshots");
int descriptor_fd = descriptor_file->get_fd();
if (descriptor_fd < 0) {
fprintf(stderr, "Unable to open descriptor output file: %m\n");
projects / cumulus.git / blobdiff