You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@quickstep.apache.org by "Marc Spehlmann (JIRA)" <ji...@apache.org> on 2016/10/31 17:15:58 UTC
[jira] [Created] (QUICKSTEP-61) SplitRowStore crashes on BulkLoads
Marc Spehlmann created QUICKSTEP-61:
---------------------------------------
Summary: SplitRowStore crashes on BulkLoads
Key: QUICKSTEP-61
URL: https://issues.apache.org/jira/browse/QUICKSTEP-61
Project: Apache Quickstep
Issue Type: Bug
Components: Storage
Environment: ubuntu 16, clang-3.8
Reporter: Marc Spehlmann
Assignee: Marc Spehlmann
Priority: Critical
When bulk loading SSB data, we noticed crashes on tables which use SplitRowStores.
On SplitRowStoreTupleStorageSubBlock.cpp:393 there is DCHECK which if you change to a CHECK should fail with some regularity when loading a new table. The surprising thing is that this behavior doesn't always happen.
You can recreate this issue sporadically by running:
```
drop table customer;
CREATE TABLE customer (
c_custkey INT NOT NULL,
c_name VARCHAR(25) NOT NULL,
c_address VARCHAR(25) NOT NULL,
c_city CHAR(10) NOT NULL,
c_nation CHAR(15) NOT NULL,
c_region CHAR(12) NOT NULL,
c_phone CHAR(15) NOT NULL,
c_mktsegment CHAR(10) NOT NULL
) WITH BLOCKPROPERTIES (
TYPE split_rowstore,
BLOCKSIZEMB 4);
copy customer from 'your_customer_tbl' with (delimiter '|');
```
Failing the check, you will get some error message like
~/incubator-quickstep/build$ ./quickstep_cli_shell --storage_path=/home/gsladmin/quickstep_store/ssb100 < q.sql
Starting Quickstep with 64 worker thread(s) and a 403.10 GB buffer pool
F1031 17:12:31.258040 91079 SplitRowStoreTupleStorageSubBlock.cpp:393] Check failed: 0 <= remaining_storage_after_inserts (0 vs. -5)
*** Check failure stack trace: ***
@ 0x3280daa (unknown)
@ 0x3283fec (unknown)
@ 0x3a276f5 (unknown)
@ 0x3a158ca (unknown)
@ 0x3a1284b (unknown)
@ 0x3a12530 (unknown)
@ 0x355f0f7 (unknown)
@ 0x3540f8c (unknown)
@ 0x331ae47 (unknown)
@ 0x329760c (unknown)
@ 0x7fdb42c51930 (unknown)
@ 0x7fdb4344e70a start_thread
@ 0x7fdb423b382d (unknown)
@ (nil) (unknown)
Aborted (core dumped)
I am currently working on this issue. Reverting #109 should resolve issues in the mean time.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)