riak_core_handoff_sender {badrecord,ho_acc}

[Konstantin74R]

Hi, please help. I can't find reson fo broken handoff between old node riak@172.16.3.92 to new node riak@172.16.3.63.

In log I see error "failed because of error:{badrecord,ho_acc}" and which node has the problem is unclear.

versions riak_kv 2.1.1 n_val = 3, nodes is 40

console.log 2023-09-13 10:48:55.880 [info] <0.25322.2>@riak_core_handoff_sender:start_fold:153 Starting ownership transfer of riak_kv_vnode from 'riak@172.16.3.92' 52808164630 1205155991565808696313200461616250880 to 'riak@172.16.3.63' 528081646301205155991565808696313200461616250880 2023-09-13 10:57:13.707 [info] <0.1490.1>@riak_kv_index_hashtree:clear_tree:842 Clearing AAE tree: 625134489405210427903529254618878869735643021312 2023-09-13 10:57:14.659 [info] <0.30711.2>@riak_kv_index_hashtree:build_or_rehash:880 Starting AAE tree build: 625134489405210427903529254618878869735643021312 2023-09-13 11:48:20.321 [error] <0.4371.0> gen_server <0.4371.0> terminated with reason: no function clause matching riak_core_handoff_sender:visit_item({<<48,98,5 8,181,35,123,106,120,146,164,42,74,122,96,253,251,232,40,185>>,<<175,18,82,141,135,102,...>>}, <<53,1,0,0,0,38,131,108,0,0,0,1,104,2,109,0,0,0,12,100,4,162,222,89, 118,233,214,0,0,43,103,104,...>>, {error,eio}) line 296 2023-09-13 11:48:20.321 [error] <0.4371.0> CRASH REPORT Process <0.4371.0> with 0 neighbours exited with reason: no function clause matching riak_core_handoff_send er:visit_item({<<48,98,58,181,35,123,106,120,146,164,42,74,122,96,253,251,232,40,185>>,<<175,18,82,141,135,102,...>>}, <<53,1,0,0,0,38,131,108,0,0,0,1,104,2,109,0, 0,0,12,100,4,162,222,89,118,233,214,0,0,43,103,104,...>>, {error,eio}) line 296 in gen_server:terminate/6 line 744 2023-09-13 11:48:20.322 [error] <0.4368.0> Supervisor {<0.4368.0>,poolboy_sup} had child riak_core_vnode_worker started with {riak_core_vnode_worker,start_link,undefined} at <0.4371.0> exit with reason no function clause matching riak_core_handoff_sender:visit_item({<<48,98,58,181,35,123,106,120,146,164,42,74,122,96,253,251,232,40,185>>,<<175,18,82,141,135,102,...>>}, <<53,1,0,0,0,38,131,108,0,0,0,1,104,2,109,0,0,0,12,100,4,162,222,89,118,233,214,0,0,43,103,104,...>>, {error,eio}) line 296 in context child_terminated 2023-09-13 11:48:20.323 [error] <0.25322.2>@riak_core_handoff_sender:start_fold:282 ownership transfer of riak_kv_vnode from 'riak@172.16.3.92' 528081646301205155991565808696313200461616250880 to 'riak@172.16.3.63' 528081646301205155991565808696313200461616250880 failed because of error:{badrecord,ho_acc} [{riak_core_handoff_sender,start_fold,5,[{file,"src/riak_core_handoff_sender.erl"},{line,204}]}] error.log

2023-09-13 10:48:48.097 [error] <0.6963.0>@riak_core_handoff_sender:start_fold:282 ownership transfer of riak_kv_vnode from 'riak@172.16.3.92' 528081646301205155991565808696313200461616250880 to 'riak@172.16.3.63' 528081646301205155991565808696313200461616250880 failed because of error:{badrecord,ho_acc} [{riak_core_handoff_sender,start_fold,5,[{file,"src/riak_core_handoff_sender.erl"},{line,204}]}] 2023-09-13 11:48:20.321 [error] <0.4371.0> gen_server <0.4371.0> terminated with reason: no function clause matching riak_core_handoff_sender:visit_item({<<48,98,58,181,35,123,106,120,146,164,42,74,122,96,253,251,232,40,185>>,<<175,18,82,141,135,102,...>>}, <<53,1,0,0,0,38,131,108,0,0,0,1,104,2,109,0,0,0,12,100,4,162,222,89,118,233,214,0,0,43,103,104,...>>, {error,eio}) line 296 2023-09-13 11:48:20.321 [error] <0.4371.0> CRASH REPORT Process <0.4371.0> with 0 neighbours exited with reason: no function clause matching riak_core_handoff_sender:visit_item({<<48,98,58,181,35,123,106,120,146,164,42,74,122,96,253,251,232,40,185>>,<<175,18,82,141,135,102,...>>}, <<53,1,0,0,0,38,131,108,0,0,0,1,104,2,109,0,0,0,12,100,4,162,222,89,118,233,214,0,0,43,103,104,...>>, {error,eio}) line 296 in gen_server:terminate/6 line 744 2023-09-13 11:48:20.322 [error] <0.4368.0> Supervisor {<0.4368.0>,poolboy_sup} had child riak_core_vnode_worker started with {riak_core_vnode_worker,start_link,undefined} at <0.4371.0> exit with reason no function clause matching riak_core_handoff_sender:visit_item({<<48,98,58,181,35,123,106,120,146,164,42,74,122,96,253,251,232,40,185>>,<<175,18,82,141,135,102,...>>}, <<53,1,0,0,0,38,131,108,0,0,0,1,104,2,109,0,0,0,12,100,4,162,222,89,118,233,214,0,0,43,103,104,...>>, {error,eio}) line 296 in context child_terminated 2023-09-13 11:48:20.323 [error] <0.25322.2>@riak_core_handoff_sender:start_fold:282 ownership transfer of riak_kv_vnode from 'riak@172.16.3.92' 528081646301205155991565808696313200461616250880 to 'riak@172.16.3.63' 528081646301205155991565808696313200461616250880 failed because of error:{badrecord,ho_acc} [{riak_core_handoff_sender,start_fold,5,[{file,"src/riak_core_handoff_sender.erl"},{line,204}]}] 2023-09-13 12:16:23.882 [error] <0.170.0> Supervisor riak_core_handoff_sender_sup had child riak_core_handoff_sender started with {riak_core_handoff_sender,start_link,undefined} at <0.2844.4> exit with reason max_concurrency in context child_terminated crash.log

2023-09-13 11:48:20 =CRASH REPORT==== crasher: initial call: riak_core_vnode_worker:init/1 pid: <0.4371.0> registered_name: [] exception exit: {{function_clause,[{riak_core_handoff_sender,visit_item,[{<<48,98,58,181,35,123,106,120,146,164,42,74,122,96,253,251,232,40,185>>,<<175,18,82,1 41,135,102,76,129,178,206,178,15,200,127,187,40,0,0,0,0>>},<<53,1,0,0,0,38,131,108,0,0,0,1,104,2,109,0,0,0,12,100,4,162,222,89,118,233,214,0,0,43,103,104,2,97,1,11 ,48,51,46,50,48,50,50,45,...>>,...],...},...]},...} ancestors: [<0.4368.0>,<0.4367.0>,<0.4366.0>,<0.1819.0>,riak_core_vnode_sup,riak_core_sup,<0.161.0>] messages: [] links: [<0.4367.0>,<0.4368.0>,<0.93.0>] dictionary: [{bitcask_file_mod,bitcask_file},{bitcask_time_fudge,no_testing},{is_visit_item_timer_set,true}] trap_exit: false status: running heap_size: 318187 stack_size: 27 reductions: 37404224106 neighbours: 2023-09-13 11:48:20 =SUPERVISOR REPORT==== Supervisor: {<0.4368.0>,poolboy_sup} Context: child_terminated Reason: {function_clause,[{riak_core_handoff_sender,visit_item,[{<<48,98,58,181,35,123,106,120,146,164,42,74,122,96,253,251,232,40,185>>,<<175,18,82,141,135,102,76,129,178,206,178,15,200,127,187,40,0,0,0,0>>},<<53,1,0,0,0,38,131,108,0,0,0,1,104,2,109,0,0,0,12,100,4,162,222,89,118,233,214,0,0,43,103,104,2,97,1,110,5, 51,46,50,48,50,50,45,...>>,...],...},...]} Offender: [{pid,<0.4371.0>},{name,riak_core_vnode_worker},{mfargs,{riak_core_vnode_worker,start_link,undefined}},{restart_type,temporary},{shutdown,5000},{child_type,worker}]

[martinsumner]

Hard to know - but the code is running a fold over the vnode and sending to the remote vnode at the time ... but in the midst of the logs is a reference to {error,eio} - a generic UNIX socket error? Is this a recurring issue, or might it have been a one-off HW or OS issue?

[Konstantin74R]

CentOS Linux release 7.9.2009 (Core) 3.10.0-1160.81.1.el7.x86_64 #1 SMP Fri Dec 16 17:29:43 UTC 2022 x86_64 x86_64 x86_64 GNU/Linux

HW and OS problems not detected.

It's a recurring problem only for one node and only for one partition. Temporary resolve - I was delete this partition from backends on this node to use replication for restore data. Joining new node to cluster was finished. No active transfers.

[martinsumner]

Perhaps there was a corrupted block on the disk within that vnode backend causing an IO error when the fold tried to read a particular part of the keyspace? Deleting the partition would then allow the handoff to complete.

You should be able to run partition repair to restore data with relative efficiency - https://docs.riak.com/riak/kv/latest/using/repair-recovery/repairs/index.html#repairing-a-single-partition.

[Konstantin74R]

I tried to start repair, but console return message to me "change ownership" and no effect. Maybe was. I not found a problems with disk.