版本更新中的ClickHouse未知编解码系列错误

Question

我一直在使用ReplicatedMergeTree.创建一个表的三节点ClickHouse集群。我最近更新了它的一个节点，从版本18.14.2到19.3.7。但是，升级后，由于数据没有同步，所以无法与升级节点建立连接。

ClickHouse错误日志如下所示。

(StorageReplicatedMergeTree): DB::StorageReplicatedMergeTree::queueTask()::<lambda(DB::StorageReplicatedMergeTree::LogEntryPtr&)>: Code: 432, e.displayText() = DB::Exception: Unknown codec family code : 130, Stack trace:

0. clickhouse-server(StackTrace::StackTrace()+0x16) [0x5d9fdf6]
1. clickhouse-server(DB::Exception::Exception(std::string const&, int)+0x1f) [0x2ceb79f]
2. clickhouse-server(DB::CompressionCodecFactory::get(unsigned char) const+0x185) [0x5af35b5]
3. clickhouse-server(DB::CompressedReadBufferBase::readCompressedData(unsigned long&, unsigned long&)+0x911) [0x5af0e31]
4. clickhouse-server(DB::CompressedReadBuffer::nextImpl()+0x26) [0x5aef7b6]
5. clickhouse-server(DB::MergeTreeDataPartChecksums::read_v3(DB::ReadBuffer&)+0x834) [0x5738604]
6. clickhouse-server(DB::MergeTreeDataPartChecksums::read_v4(DB::ReadBuffer&)+0xb2) [0x5738fd2]
7. clickhouse-server(DB::MergeTreeDataPartChecksums::read(DB::ReadBuffer&, unsigned long)+0x28) [0x5739f68]
8. clickhouse-server(DB::MergeTreeDataPartChecksums::read(DB::ReadBuffer&)+0x94) [0x573a114]
9. clickhouse-server(DB::MergeTreeDataPart::loadChecksums(bool)+0x119) [0x572f6b9]
10. clickhouse-server(DB::MergeTreeDataPart::loadColumnsChecksumsIndexes(bool, bool)+0x39) [0x57338a9]
11. clickhouse-server(DB::DataPartsExchange::Fetcher::fetchPart(std::string const&, std::string const&, std::string const&, int, DB::ConnectionTimeouts const&, std::string const&, std::string const&, std::string const&, bool, std::string const&)+0xdde) [0x56d955e]
12. clickhouse-server() [0x567a556]
13. clickhouse-server(DB::StorageReplicatedMergeTree::fetchPart(std::string const&, std::string const&, bool, unsigned long)+0x7a9) [0x569c379]
14. clickhouse-server(DB::StorageReplicatedMergeTree::executeFetch(DB::ReplicatedMergeTreeLogEntry&)+0x999) [0x569ebf9]
15. clickhouse-server(DB::StorageReplicatedMergeTree::executeLogEntry(DB::ReplicatedMergeTreeLogEntry&)+0x9f) [0x56a3c9f]
16. clickhouse-server() [0x56a4610]
17. clickhouse-server(DB::ReplicatedMergeTreeQueue::processEntry(std::function<std::shared_ptr<zkutil::ZooKeeper> ()>, std::shared_ptr<DB::ReplicatedMergeTreeLogEntry>&, std::function<bool (std::shared_ptr<DB::ReplicatedMergeTreeLogEntry>&)>)+0x3e) [0x57aadfe]
18. clickhouse-server(DB::StorageReplicatedMergeTree::queueTask()+0x192) [0x5678b62]
19. clickhouse-server(DB::BackgroundProcessingPool::threadFunction()+0x396) [0x56d3b86]
20. clickhouse-server() [0x56d4484]
21. clickhouse-server(ThreadPoolImpl<std::thread>::worker(std::_List_iterator<std::thread>)+0x199) [0x5da8aa9]
22. clickhouse-server() [0x65ce86f]
23. /lib64/libpthread.so.0(+0x7e25) [0x7fb7b43ffe25]
24. /lib64/libc.so.6(clone+0x6d) [0x7fb7b3c2334d]

编辑1:

升级到19.4.3.11修正了此错误。但是，现在继续获取以下错误，但仍无法与服务器建立连接。

(StorageReplicatedMergeTree)：DB::StorageReplicatedMergeTree::queueTask()::：Poco::Exception。代码: 1000，文件"/home/centos/clickhouse-rpm/rpmbuild/BUILD/ClickHouse-19.4.3.11-stable/contrib/poco/Util/include/Poco/Util/Application.h"，第457行中的e.code() = 0，e.displayText() = Null指针：_pInstance

编辑2:

发现上述是ClickHouse v19.4.3.11 - https://github.com/yandex/ClickHouse/issues/5005中的另一个bug

Answer 1

我遇到了同样的异常，需要对失败的节点执行下面的查询(请参阅github: DB：：异常:未知编解码器家族代码:4)：

SYSTEM DROP MARK CACHE