TiFlash节点不停重启

【 TiDB 使用环境】生产环境 【 TiDB 版本】5.0.6 【遇到的问题】TiFlash节点不停重启 【复现路径】做过哪些操作出现的问题` 【问题现象及影响】 Dashboard上看一个TiFlash节点离线,登录节点看隔一阵子就重启,也没有数据同步。 【附件】 相关日志及监控(https://metricstool.pingcap.com/) tiflash_error.log image

若提问为性能优化、故障排查类问题,请下载脚本运行。终端输出的打印结果,请务必全选并复制粘贴上传。

2022.07.25 16:01:53.394712 [ 1 ] Application: The configuration “path” is deprecated. Check [storage] section for new style. 2022.07.25 16:25:50.967089 [ 28 ] pingcap.tikv: region {295876070,1907,3138} find error: region 295876070 is missing 2022.07.25 16:25:53.829955 [ 38 ] void DB::ApplyPreHandledSnapshot(DB::EngineStoreServerWrap*, DB::PreHandledSnapshot*): Code: 49, e.displayText() = DB::Exception: DB::KVStore::checkAndApplySnapshot(const DB::RegionPtrWithBlock&, DB::TMTContext&)::<lambda(DB::RegionMap, const DB::KVStoreTaskLock&)>: range of region 270272686 is overlapped with region 172787948, should not happen, e.what() = DB::Exception, Stack t race:

  1. bin/tiflash/tiflash(StackTrace::StackTrace()+0x15) [0x36c09e5]
  2. bin/tiflash/tiflash(DB::Exception::Exception(std::__cxx11::basic_string<char, std::char_traits, std::allocator > const&, int)+0x25) [0x36b7575]
  3. bin/tiflash/tiflash() [0x7710255]
  4. bin/tiflash/tiflash(DB::KVStore::handleRegionsByRangeOverlap(std::pair<DB::TiKVRangeKey, DB::TiKVRangeKey> const&, std::function<void (std::unordered_map<unsigned long, std::shared_ptrDB::Region, std::hash, std::equal_to, std::allocator<std::pair<unsigned long const, std::shared_ptrDB::Region > > >, DB::KVStoreTaskLock const&)>&&) const+0x52) [0x7314d92]
  5. bin/tiflash/tiflash(DB::KVStore::checkAndApplySnapshot(DB::RegionPtrWithBlock const&, DB::TMTContext&)+0x291) [0x7711441]
  6. bin/tiflash/tiflash(DB::KVStore::handlePreApplySnapshot(DB::RegionPtrWithBlock const&, DB::TMTContext&)+0x162) [0x77125e2]
  7. bin/tiflash/tiflash(DB::ApplyPreHandledSnapshot(DB::EngineStoreServerWrap*, DB::PreHandledSnapshot*)+0x3f) [0x732497f]
  8. bin/tiflash/tiflash(DB::ApplyPreHandledSnapshot(DB::EngineStoreServerWrap*, void*, unsigned int)+0x1d) [0x7324a3d]
  9. /chj/app/tidb/deploy/tiflash-9000/bin/tiflash/libtiflash_proxy.so(+0xc93e59) [0x7fe99301ce59]
  10. /chj/app/tidb/deploy/tiflash-9000/bin/tiflash/libtiflash_proxy.so(+0xc8eaf2) [0x7fe993017af2]
  11. /chj/app/tidb/deploy/tiflash-9000/bin/tiflash/libtiflash_proxy.so(+0x9fb326) [0x7fe992d84326]
  12. /chj/app/tidb/deploy/tiflash-9000/bin/tiflash/libtiflash_proxy.so(+0x1db7640) [0x7fe994140640]
  13. /chj/app/tidb/deploy/tiflash-9000/bin/tiflash/libtiflash_proxy.so(+0x563206) [0x7fe9928ec206]
  14. /chj/app/tidb/deploy/tiflash-9000/bin/tiflash/libtiflash_proxy.so(+0x59d9ff) [0x7fe9929269ff]
  15. /chj/app/tidb/deploy/tiflash-9000/bin/tiflash/libtiflash_proxy.so(+0x1577c9f) [0x7fe993900c9f]
  16. /lib64/libpthread.so.0(+0x7e64) [0x7fe991b4ee64]
  17. /lib64/libc.so.6(clone+0x6c) [0x7fe99157588c]

2022.07.25 16:26:11.648538 [ 1 ] Application: The configuration “path” is deprecated. Check [storage] section for new style. 2022.07.25 16:51:01.082572 [ 32 ] pingcap.tikv: region {295507460,1913,2785} find error: EpochNotMatch current epoch of region 295507460 is conf_ver: 1913 version: 2786, but you sent conf_ver: 1913 vers ion: 2785 2022.07.25 16:51:02.979909 [ 32 ] pingcap.tikv: region {295724190,1907,3140} find error: EpochNotMatch current epoch of region 295724190 is conf_ver: 1907 version: 3141, but you sent conf_ver: 1907 vers ion: 3140 2022.07.25 16:51:05.811363 [ 38 ] void DB::ApplyPreHandledSnapshot(DB::EngineStoreServerWrap*, DB::PreHandledSnapshot*): Code: 49, e.displayText() = DB::Exception: DB::KVStore::checkAndApplySnapshot(const DB::RegionPtrWithBlock&, DB::TMTContext&)::<lambda(DB::RegionMap, const DB::KVStoreTaskLock&)>: range of region 270272686 is overlapped with region 172787948, should not happen, e.what() = DB::Exception, Stack t race:

  1. bin/tiflash/tiflash(StackTrace::StackTrace()+0x15) [0x36c09e5]
  2. bin/tiflash/tiflash(DB::Exception::Exception(std::__cxx11::basic_string<char, std::char_traits, std::allocator > const&, int)+0x25) [0x36b7575]
  3. bin/tiflash/tiflash() [0x7710255]
  4. bin/tiflash/tiflash(DB::KVStore::handleRegionsByRangeOverlap(std::pair<DB::TiKVRangeKey, DB::TiKVRangeKey> const&, std::function<void (std::unordered_map<unsigned long, std::shared_ptrDB::Region, std::hash, std::equal_to, std::allocator<std::pair<unsigned long const, std::shared_ptrDB::Region > > >, DB::KVStoreTaskLock const&)>&&) const+0x52) [0x7314d92]
  5. bin/tiflash/tiflash(DB::KVStore::checkAndApplySnapshot(DB::RegionPtrWithBlock const&, DB::TMTContext&)+0x291) [0x7711441]
  6. bin/tiflash/tiflash(DB::KVStore::handlePreApplySnapshot(DB::RegionPtrWithBlock const&, DB::TMTContext&)+0x162) [0x77125e2]
  7. bin/tiflash/tiflash(DB::ApplyPreHandledSnapshot(DB::EngineStoreServerWrap*, DB::PreHandledSnapshot*)+0x3f) [0x732497f]
  8. bin/tiflash/tiflash(DB::ApplyPreHandledSnapshot(DB::EngineStoreServerWrap*, void*, unsigned int)+0x1d) [0x7324a3d]
  9. /chj/app/tidb/deploy/tiflash-9000/bin/tiflash/libtiflash_proxy.so(+0xc93e59) [0x7f66ef747e59]
  10. /chj/app/tidb/deploy/tiflash-9000/bin/tiflash/libtiflash_proxy.so(+0xc8eaf2) [0x7f66ef742af2]
  11. /chj/app/tidb/deploy/tiflash-9000/bin/tiflash/libtiflash_proxy.so(+0x9fb326) [0x7f66ef4af326]
  12. /chj/app/tidb/deploy/tiflash-9000/bin/tiflash/libtiflash_proxy.so(+0x1db7640) [0x7f66f086b640]
  13. /chj/app/tidb/deploy/tiflash-9000/bin/tiflash/libtiflash_proxy.so(+0x563206) [0x7f66ef017206]
  14. /chj/app/tidb/deploy/tiflash-9000/bin/tiflash/libtiflash_proxy.so(+0x59d9ff) [0x7f66ef0519ff]
  15. /chj/app/tidb/deploy/tiflash-9000/bin/tiflash/libtiflash_proxy.so(+0x1577c9f) [0x7f66f002bc9f]
  16. /lib64/libpthread.so.0(+0x7e64) [0x7f66ee279e64]
  17. /lib64/libc.so.6(clone+0x6c) [0x7f66edca088c]

2022.07.25 16:51:23.645341 [ 1 ] Application: The configuration “path” is deprecated. Check [storage] section for new style. 2022.07.25 17:16:28.651135 [ 38 ] void DB::ApplyPreHandledSnapshot(DB::EngineStoreServerWrap*, DB::PreHandledSnapshot*): Code: 49, e.displayText() = DB::Exception: DB::KVStore::checkAndApplySnapshot(const DB::RegionPtrWithBlock&, DB::TMTContext&)::<lambda(DB::RegionMap, const DB::KVStoreTaskLock&)>: range of region 270272686 is overlapped with region 172787948, should not happen, e.what() = DB::Exception, Stack t race:

  1. bin/tiflash/tiflash(StackTrace::StackTrace()+0x15) [0x36c09e5]
  2. bin/tiflash/tiflash(DB::Exception::Exception(std::__cxx11::basic_string<char, std::char_traits, std::allocator > const&, int)+0x25) [0x36b7575]
  3. bin/tiflash/tiflash() [0x7710255]
  4. bin/tiflash/tiflash(DB::KVStore::handleRegionsByRangeOverlap(std::pair<DB::TiKVRangeKey, DB::TiKVRangeKey> const&, std::function<void (std::unordered_map<unsigned long, std::shared_ptrDB::Region, std::hash, std::equal_to, std::allocator<std::pair<unsigned long const, std::shared_ptrDB::Region > > >, DB::KVStoreTaskLock const&)>&&) const+0x52) [0x7314d92]
  5. bin/tiflash/tiflash(DB::KVStore::checkAndApplySnapshot(DB::RegionPtrWithBlock const&, DB::TMTContext&)+0x291) [0x7711441]
  6. bin/tiflash/tiflash(DB::KVStore::handlePreApplySnapshot(DB::RegionPtrWithBlock const&, DB::TMTContext&)+0x162) [0x77125e2]
  7. bin/tiflash/tiflash(DB::ApplyPreHandledSnapshot(DB::EngineStoreServerWrap*, DB::PreHandledSnapshot*)+0x3f) [0x732497f]
  8. bin/tiflash/tiflash(DB::ApplyPreHandledSnapshot(DB::EngineStoreServerWrap*, void*, unsigned int)+0x1d) [0x7324a3d]
  9. /chj/app/tidb/deploy/tiflash-9000/bin/tiflash/libtiflash_proxy.so(+0xc93e59) [0x7fbf58a65e59]
  10. /chj/app/tidb/deploy/tiflash-9000/bin/tiflash/libtiflash_proxy.so(+0xc8eaf2) [0x7fbf58a60af2]
  11. /chj/app/tidb/deploy/tiflash-9000/bin/tiflash/libtiflash_proxy.so(+0x9fb326) [0x7fbf587cd326]
  12. /chj/app/tidb/deploy/tiflash-9000/bin/tiflash/libtiflash_proxy.so(+0x1db7640) [0x7fbf59b89640]
  13. /chj/app/tidb/deploy/tiflash-9000/bin/tiflash/libtiflash_proxy.so(+0x563206) [0x7fbf58335206]
  14. /chj/app/tidb/deploy/tiflash-9000/bin/tiflash/libtiflash_proxy.so(+0x59d9ff) [0x7fbf5836f9ff]
  15. /chj/app/tidb/deploy/tiflash-9000/bin/tiflash/libtiflash_proxy.so(+0x1577c9f) [0x7fbf59349c9f]
  16. /lib64/libpthread.so.0(+0x7e64) [0x7fbf57597e64]
  17. /lib64/libc.so.6(clone+0x6c) [0x7fbf56fbe88c]

2022.07.25 17:16:46.643977 [ 1 ] Application: The configuration “path” is deprecated. Check [storage] section for new style.

感觉和这个有点像呀

暂时不能重启有好的办法么

尝试重新缩容、扩容后出现了节点无法去除信息的情况。

https://github.com/pingcap/tiflash/issues/3435 升级

暂时还不能升级,现在是这个情况https://asktug.com/t/topic/813169