mirror/ternfs-XTXMarkets
1
0
Fork 0
mirror of https://github.com/XTXMarkets/ternfs.git synced 2026-05-05 01:09:35 -05:00
Code Issues Packages Projects Releases Wiki Activity

eggsdbtools: fix nameHash mismatch in edge keys for shard 0

Browse Source
This commit is contained in:
Miroslav Crnic
2024-07-29 16:08:28 +01:00
committed by GitHub Enterprise
parent cd286fcd81
commit 95a2681121
6 changed files with 103 additions and 37 deletions
Download Patch File Download Diff File Expand all files Collapse all files
cpp/dbtools/ShardDBTools.cpp
+76 -17
View File
@@ -4,8 +4,11 @@
#include <memory>
#include <rocksdb/db.h>
#include <rocksdb/slice.h>
#include <rocksdb/write_batch.h>
#include <vector>
#include "Bincode.hpp"
#include "BlockServicesCacheDB.hpp"
#include "Common.hpp"
#include "Env.hpp"
#include "LogsDB.hpp"
@@ -178,14 +181,6 @@ void ShardDBTools::fsck(const std::string& dbPath) {
LOG_INFO(env, "Analyzed %s directories", analyzedDirs);
}
// no duplicate ownership of any kind in edges -- currently broken in shard 0
// because of the outage which lead to the fork
bool checkEdgeOrdering = dbPath.find("000") == std::string::npos;
if (!checkEdgeOrdering) {
LOG_INFO(env, "Will not check edge ordering since we think this is shard 0");
}
{
LOG_INFO(env, "Edges pass");
const rocksdb::ReadOptions options;
@@ -234,7 +229,11 @@ void ShardDBTools::fsck(const std::string& dbPath) {
thisDirHasCurrentEdges = false;
thisDirMaxTime = 0;
}
auto nameHash = edgeK().nameHash();
std::string name(edgeK().name().data(), edgeK().name().size());
if (nameHash != EdgeKey::computeNameHash(HashMode::XXH3_63, BincodeBytesRef(edgeK().name().data(), edgeK().name().size()))) {
ERROR("Edge %s has mismatch between name and nameHash", edgeK());
}
InodeId ownedTargetId = NULL_INODE_ID;
EggsTime creationTime;
std::optional<ExternalValue<CurrentEdgeBody>> currentEdge;
@@ -296,23 +295,17 @@ void ShardDBTools::fsck(const std::string& dbPath) {
(prevCreationTime > creationTime) ||
(prevCreationTime == creationTime && !(snapshotEdge && (*snapshotEdge)().targetIdWithOwned().id() == NULL_INODE_ID))
) {
if (checkEdgeOrdering) {
ERROR("Name %s in directory %s has overlapping edges (%s >= %s).", name, edgeK().dirId(), prevCreationTime, creationTime);
}
ERROR("Name %s in directory %s has overlapping edges (%s >= %s).", name, edgeK().dirId(), prevCreationTime, creationTime);
}
// Deletion edges are not preceded by another deletion edge
if (snapshotEdge && (*snapshotEdge)().targetIdWithOwned().id() == NULL_INODE_ID && prevEdge->second.second().targetIdWithOwned().id() == NULL_INODE_ID) {
if (checkEdgeOrdering) {
ERROR("Deletion edge with name %s creation time %s is preceded by another deletion edge in directory %s", name, edgeK().creationTime(), thisDir);
}
ERROR("Deletion edge with name %s creation time %s is preceded by another deletion edge in directory %s", name, edgeK().creationTime(), thisDir);
}
} else {
// Deletion edges are preceded by something not -- this is actually not enforced
// by the schema, but by GC. So it's not a huge deal if it's wrong.
if (snapshotEdge && (*snapshotEdge)().targetIdWithOwned().id() == NULL_INODE_ID) {
if (checkEdgeOrdering) {
ERROR("Deletion edge with name %s creation time %s is not preceded by anything in directory %s", name, edgeK().creationTime(), thisDir);
}
ERROR("Deletion edge with name %s creation time %s is not preceded by anything in directory %s", name, edgeK().creationTime(), thisDir);
}
}
}
@@ -452,3 +445,69 @@ void ShardDBTools::fsck(const std::string& dbPath) {
#undef ERROR
}
void ShardDBTools::fixNameHashMismatch(const std::string& dbPath) {
Logger logger(LogLevel::LOG_INFO, STDERR_FILENO, false, false);
std::shared_ptr<XmonAgent> xmon;
Env env(logger, xmon, "ShardDBTools");
SharedRocksDB sharedDb(logger, xmon, dbPath, "");
sharedDb.registerCFDescriptors(ShardDB::getColumnFamilyDescriptors());
sharedDb.registerCFDescriptors(LogsDB::getColumnFamilyDescriptors());
sharedDb.registerCFDescriptors(BlockServicesCacheDB::getColumnFamilyDescriptors());
rocksdb::Options rocksDBOptions;
rocksDBOptions.compression = rocksdb::kLZ4Compression;
rocksDBOptions.bottommost_compression = rocksdb::kZSTD;
sharedDb.open(rocksDBOptions);
auto db = sharedDb.db();
{
auto edgesCf = sharedDb.getCF("edges");
const rocksdb::ReadOptions options;
std::vector<std::string> keysToDelete;
std::vector<std::pair<std::string,std::string>> keyValuesToInsert;
std::unique_ptr<rocksdb::Iterator> it(db->NewIterator(options, edgesCf));
for (it->SeekToFirst(); it->Valid(); it->Next()) {
auto edgeK = ExternalValue<EdgeKey>::FromSlice(it->key());
if (!edgeK().current()) {
continue;
}
auto nameHash = edgeK().nameHash();
std::string name(edgeK().name().data(), edgeK().name().size());
auto computedHash = EdgeKey::computeNameHash(HashMode::XXH3_63, BincodeBytesRef(edgeK().name().data(), edgeK().name().size()));
if (nameHash == computedHash) {
continue;
}
const std::string suffix = ".resurrected";
if (name.compare(name.size() - suffix.size(), suffix.size(), suffix) != 0) {
LOG_ERROR(env, "found an edge with name hash mismatch not ending in '.ressurected': %s", edgeK());
continue;
}
keysToDelete.emplace_back(it->key().ToString());
StaticValue<EdgeKey> newEdgeKey;
newEdgeKey().setDirIdWithCurrentU64(edgeK().dirIdWithCurrentU64());
newEdgeKey().setName(edgeK().name());
newEdgeKey().setNameHash(computedHash);
keyValuesToInsert.emplace_back(newEdgeKey.toSlice().ToString(),it->value().ToString());
}
ROCKS_DB_CHECKED(it->status());
if (keysToDelete.size() != keyValuesToInsert.size()) {
LOG_ERROR(env, "bug keys to delete different than keys to insert");
return;
}
if (keysToDelete.size() != 105) {
LOG_ERROR(env, "unexpected number of keys to delete %s", keysToDelete.size());
return;
}
rocksdb::WriteBatch batch;
for (const auto& key : keysToDelete) {
batch.Delete(edgesCf, key);
}
for (const auto& [key, value] : keyValuesToInsert) {
batch.Put(edgesCf, key, value);
}
ROCKS_DB_CHECKED(db->Write(rocksdb::WriteOptions(), &batch));
}
sharedDb.close();
}