SlideShare a Scribd company logo
Eat My Data: (now with 20% more rant!) How everybody gets file I/O wrong Stewart Smith [email_address] Senior Software Engineer, MySQL Cluster MySQL AB
What I work on ,[object Object],[object Object],[object Object],[object Object]
Overview ,[object Object]
Overview ,[object Object],[object Object],[object Object]
Overview ,[object Object],[object Object],[object Object],[object Object]
Overview ,[object Object],[object Object],[object Object],[object Object],[object Object]
Overview ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
In the beginning ,[object Object]
In the beginning ,[object Object],[object Object]
In the beginning ,[object Object],[object Object],[object Object]
A world without failure
A world without failure ,[object Object]
A world without failure ,[object Object],[object Object]
A world without failure ,[object Object],[object Object],[object Object]
A world without failure ,[object Object],[object Object],[object Object],[object Object]
A world without failure ,[object Object],[object Object],[object Object],[object Object],[object Object]
A world without failure ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
A world without failure ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Data Consistency ,[object Object]
User Expectations ,[object Object]
User Expectations ,[object Object],[object Object]
User Expectations ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Databases
Databases ,[object Object]
Databases ,[object Object],[object Object]
Databases ,[object Object],[object Object],[object Object],[object Object]
Databases ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Databases ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Databases ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Databases ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
What databases are good at ,[object Object]
What databases are good at ,[object Object],[object Object]
What databases are good at ,[object Object],[object Object],[object Object]
What databases are good at ,[object Object],[object Object],[object Object],[object Object],[object Object]
Easy solution to data consistency ,[object Object],[object Object],[object Object]
Revelation #1 ,[object Object]
Revelation #2 ,[object Object]
Revelation #3 ,[object Object]
Revelation #3 ,[object Object],[object Object]
Eat my data ,[object Object]
Where data can be ,[object Object],[object Object]
Where data can be ,[object Object],[object Object],[object Object],[object Object]
Where data can be ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Where data can be ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Data flow ,[object Object],[object Object]
Data flow ,[object Object],[object Object],[object Object],[object Object]
Data flow ,[object Object],[object Object],[object Object],[object Object],[object Object]
Data flow ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Data flow ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Data flow ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Data flow ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Simple Application: Save==on disk ,[object Object],[object Object],[object Object]
Saving a simple document ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Bug #1 ,[object Object],[object Object]
Word Processor Saving -1 Bug ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Bug #2, 3 and 4 ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
File System Integrity ,[object Object]
File System Integrity ,[object Object],[object Object]
File System Integrity ,[object Object],[object Object],[object Object]
File System Integrity ,[object Object],[object Object],[object Object],[object Object]
File System Integrity ,[object Object],[object Object],[object Object],[object Object],[object Object]
Data journaling ,[object Object]
Atomic write(2)
Atomic write(2) ,[object Object]
Atomic write(2) ,[object Object],[object Object]
Atomic write(2) ,[object Object],[object Object],[object Object]
Atomic write(2) ,[object Object],[object Object],[object Object],[object Object]
Atomic write(2) ,[object Object],[object Object],[object Object],[object Object],[object Object]
Eat My Data ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],CRASH
Write to Temp file, rename ,[object Object]
Write to Temp file, rename ,[object Object],[object Object],[object Object]
Write to Temp file, rename ,[object Object],[object Object],[object Object],[object Object]
Write to Temp file, rename ,[object Object],[object Object],[object Object],[object Object],[object Object]
Write to Temp file, rename ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Temp file, rename ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Now all is good with the world...
Now all is good with the world... ,[object Object]
Now all is good with the world... ,[object Object],[object Object]
[object Object]
Now all is good with the world... ,[object Object],[object Object],[object Object]
Now all is good with the world... ,[object Object],[object Object],[object Object],[object Object]
File System Integrity ,[object Object],[object Object],[object Object]
File System Integrity ,[object Object],[object Object],[object Object],[object Object]
File System Integrity ,[object Object],[object Object],[object Object],[object Object],[object Object]
data=ordered ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
other systems ,[object Object],[object Object],[object Object],[object Object],[object Object]
flush and sync ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],Flush the buffers! Sync to disk  before  rename
A tale of libxml2 ,[object Object]
A tale of libxml2 ,[object Object],[object Object]
A tale of libxml2 ,[object Object],[object Object],[object Object]
A tale of libxml2 ,[object Object],[object Object],[object Object],[object Object]
A tale of libxml2 ,[object Object],[object Object],[object Object],[object Object],[object Object]
so, replace ,[object Object]
gint common_save_xml(xmlDocPtr doc, gchar *filename) { FILE  *fp; char  *xmlbuf; int  fd, n; fp = g_fopen(filename, &quot;w&quot;); if(NULL == fp) return -1; xmlDocDumpFormatMemory(doc, (xmlChar **)&xmlbuf, &n, TRUE); if(n <= 0) { errno = ENOMEM; return -1; } if(fwrite(xmlbuf, sizeof (xmlChar), n, fp) < n) { xmlFree (xmlbuf); return -1; } xmlFree (xmlbuf); /* flush user-space buffers */ if (fflush (fp) != 0) return -1; if ((fd = fileno (fp)) == -1) return -1; #ifdef HAVE_FSYNC /* sync kernel-space buffers to disk */ if (fsync (fd) == -1) return -1; #endif fclose(fp); return 0; }
Nearing Nirvana ,[object Object],[object Object]
Except if you want to be portable... ,[object Object],[object Object]
Except if you want to be portable... ,[object Object],[object Object],[object Object]
Except if you want to be portable... ,[object Object],[object Object],[object Object]
on fsync, POSIX Says... ,[object Object]
on fsync, POSIX Says... ,[object Object]
POSIX compliant fsync ,[object Object]
POSIX compliant fsync ,[object Object],[object Object]
POSIX compliant fsync ,[object Object],[object Object],[object Object]
POSIX compliant fsync ,[object Object],[object Object],[object Object],[object Object]
POSIX compliant fsync ,[object Object],[object Object],[object Object],[object Object],gcc
POSIX compliant fsync ,[object Object],[object Object],[object Object],[object Object],pushl  %ebp movl  %esp, %ebp movl  $0, %eax popl  %ebp ret gcc
Tale of a really fast database server ,[object Object]
Tale of a really fast database server ,[object Object],[object Object]
Tale of a really fast database server ,[object Object],[object Object],[object Object]
Tale of a really fast database server ,[object Object],[object Object],[object Object],[object Object]
fsync() doesn't have to sync ,[object Object]
fsync() doesn't have to sync ,[object Object],[object Object]
Standards are great ,[object Object]
Standards are great ,[object Object],[object Object]
Standards are great ,[object Object],[object Object],[object Object]
Standards are great ,[object Object],[object Object],[object Object],[object Object]
#ifdef HAVE_DARWIN_THREADS # ifdef F_FULLFSYNC /* This executable has been compiled on Mac OS X 10.3 or later. Assume that F_FULLFSYNC is available at run-time. */ srv_have_fullfsync = TRUE; # else /* F_FULLFSYNC */ /* This executable has been compiled on Mac OS X 10.2 or earlier.  Determine if the executable is running on Mac OS X 10.3 or later. */ struct utsname utsname; if (uname(&utsname)) { fputs(&quot;InnoDB: cannot determine Mac OS X version!&quot;, stderr); } else { srv_have_fullfsync = strcmp(utsname.release, &quot;7.&quot;) >= 0; } if (!srv_have_fullfsync) { fputs(&quot;InnoDB: On Mac OS X, fsync() may be&quot; &quot; broken on internal drives,&quot; &quot;InnoDB: making transactions unsafe!&quot;, stderr); } # endif /* F_FULLFSYNC */ #endif /* HAVE_DARWIN_THREADS */
#if defined(HAVE_DARWIN_THREADS) # ifndef F_FULLFSYNC /* The following definition is from the Mac OS X 10.3 <sys/fcntl.h> */ #  define F_FULLFSYNC 51 /* fsync + ask the drive to flush to the media */ # elif F_FULLFSYNC != 51 #  error &quot;F_FULLFSYNC != 51: ABI incompatibility with Mac OS X 10.3&quot; # endif /* Apple has disabled fsync() for internal disk drives in OS X. That caused corruption for a user when he tested a power outage. Let us in OS X use a nonstandard flush method recommended by an Apple engineer. */ if (!srv_have_fullfsync) { /* If we are not on an operating system that supports this, then fall back to a plain fsync. */ ret = fsync(file); } else { ret = fcntl(file, F_FULLFSYNC, NULL); if (ret) { /* If we are not on a file system that supports this, then fall back to a plain fsync. */ ret = fsync(file); } } #elif HAVE_FDATASYNC ret = fdatasync(file); #else /*  fprintf(stderr, &quot;Flushing to file %p&quot;, file); */ ret = fsync(file); #endif
Yes, some OS Vendors hate you ,[object Object]
Big Files
Big Files ,[object Object]
Big Files ,[object Object],[object Object]
Big Files ,[object Object],[object Object],[object Object]
Big Files ,[object Object],[object Object],[object Object],[object Object]
Big Files ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Big Files ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Large directories ,[object Object]
Large directories ,[object Object],[object Object]
Large directories ,[object Object],[object Object],[object Object]
Large directories ,[object Object],[object Object],[object Object],[object Object],[object Object]
Large directories ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Where data is after being written ,[object Object]
Where data is after being written ,[object Object],[object Object],[object Object]
Where data is after being written ,[object Object],[object Object],[object Object],[object Object]
Where data is after being written ,[object Object],[object Object],[object Object],[object Object],[object Object]
Where data is after being written ,[object Object],[object Object],[object Object],[object Object],[object Object]
Where data is after being written ,[object Object],[object Object],[object Object],[object Object],[object Object]
sqlite ,[object Object]
sqlite ,[object Object],[object Object]
sqlite ,[object Object],[object Object],[object Object]
sqlite ,[object Object],[object Object],[object Object],[object Object]
sqlite ,[object Object],[object Object],[object Object],[object Object],[object Object]
sqlite ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
sqlite ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Performance of Large files ,[object Object],[object Object]
Performance of Large files ,[object Object],[object Object],[object Object]
Performance of Large files ,[object Object],[object Object],[object Object],[object Object]
Performance of Large files ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
inode Infos Direct blocks Indirect Blocks Double indirect blocks
Extent ,[object Object],[object Object],[object Object],[object Object],[object Object]
Parallel writers ,[object Object]
Parallel writers ,[object Object],[object Object],[object Object]
Parallel writers ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Parallel writers ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Preallocation ,[object Object],[object Object],[object Object],[object Object],[object Object]
Tablespace allocation in NDB ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Improvements in mysql-test-run ,[object Object],[object Object],[object Object]
Improvements in mysql-test-run ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Improvements in mysql-test-run ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Library Developers ,[object Object]
Library Developers ,[object Object],[object Object]
Library Developers ,[object Object],[object Object],[object Object]
Library Developers ,[object Object],[object Object],[object Object],[object Object]
Library Developers ,[object Object],[object Object],[object Object],[object Object],[object Object]
Library Developers ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
There is hope ,[object Object],[object Object],[object Object]
Good Luck!
Good Luck! ,[object Object]

More Related Content

What's hot

ProdSec: A Technical Approach
ProdSec: A Technical ApproachProdSec: A Technical Approach
ProdSec: A Technical Approach
Jeremy Brown
 
Datadog- Monitoring In Motion
Datadog- Monitoring In Motion Datadog- Monitoring In Motion
Datadog- Monitoring In Motion
Cloud Native Apps SF
 
Distributed Counters in Cassandra (Cassandra Summit 2010)
Distributed Counters in Cassandra (Cassandra Summit 2010)Distributed Counters in Cassandra (Cassandra Summit 2010)
Distributed Counters in Cassandra (Cassandra Summit 2010)
kakugawa
 
Fun with the Hak5 Rubber Ducky
Fun with the Hak5 Rubber DuckyFun with the Hak5 Rubber Ducky
Fun with the Hak5 Rubber Ducky
kieranjacobsen
 
Sticky Keys to the Kingdom
Sticky Keys to the KingdomSticky Keys to the Kingdom
Sticky Keys to the Kingdom
Dennis Maldonado
 
A Brief History of Cryptographic Failures
A Brief History of Cryptographic FailuresA Brief History of Cryptographic Failures
A Brief History of Cryptographic Failures
Nothing Nowhere
 
MongoDB World 2019: MongoDB Read Isolation: Making Your Reads Clean, Committe...
MongoDB World 2019: MongoDB Read Isolation: Making Your Reads Clean, Committe...MongoDB World 2019: MongoDB Read Isolation: Making Your Reads Clean, Committe...
MongoDB World 2019: MongoDB Read Isolation: Making Your Reads Clean, Committe...
MongoDB
 
NameNode Analytics - Querying HDFS Namespace in Real Time
NameNode Analytics - Querying HDFS Namespace in Real TimeNameNode Analytics - Querying HDFS Namespace in Real Time
NameNode Analytics - Querying HDFS Namespace in Real Time
Plamen Jeliazkov
 
CNIT 121: 17 Remediation Introduction (Part 1)
CNIT 121: 17 Remediation Introduction (Part 1)CNIT 121: 17 Remediation Introduction (Part 1)
CNIT 121: 17 Remediation Introduction (Part 1)
Sam Bowne
 
No Easy Breach DerbyCon 2016
No Easy Breach DerbyCon 2016No Easy Breach DerbyCon 2016
No Easy Breach DerbyCon 2016
Matthew Dunwoody
 
EDB Postgres DBA Best Practices
EDB Postgres DBA Best PracticesEDB Postgres DBA Best Practices
EDB Postgres DBA Best Practices
EDB
 
WAF Bypass Techniques - Using HTTP Standard and Web Servers’ Behaviour
WAF Bypass Techniques - Using HTTP Standard and Web Servers’ BehaviourWAF Bypass Techniques - Using HTTP Standard and Web Servers’ Behaviour
WAF Bypass Techniques - Using HTTP Standard and Web Servers’ Behaviour
Soroush Dalili
 
Domino Server Health - Monitoring and Managing
 Domino Server Health - Monitoring and Managing Domino Server Health - Monitoring and Managing
Domino Server Health - Monitoring and Managing
Gabriella Davis
 
C* Summit 2013: The World's Next Top Data Model by Patrick McFadin
C* Summit 2013: The World's Next Top Data Model by Patrick McFadinC* Summit 2013: The World's Next Top Data Model by Patrick McFadin
C* Summit 2013: The World's Next Top Data Model by Patrick McFadin
DataStax Academy
 
lotus domino questions answers 2017
lotus domino questions answers 2017lotus domino questions answers 2017
lotus domino questions answers 2017
mJOBrr
 
Architecture Sustaining LINE Sticker services
Architecture Sustaining LINE Sticker servicesArchitecture Sustaining LINE Sticker services
Architecture Sustaining LINE Sticker services
LINE Corporation
 
Extending Flink State Serialization for Better Performance and Smaller Checkp...
Extending Flink State Serialization for Better Performance and Smaller Checkp...Extending Flink State Serialization for Better Performance and Smaller Checkp...
Extending Flink State Serialization for Better Performance and Smaller Checkp...
Flink Forward
 
HTTP Request Smuggling via higher HTTP versions
HTTP Request Smuggling via higher HTTP versionsHTTP Request Smuggling via higher HTTP versions
HTTP Request Smuggling via higher HTTP versions
neexemil
 
The Volcano/Cascades Optimizer
The Volcano/Cascades OptimizerThe Volcano/Cascades Optimizer
The Volcano/Cascades Optimizer
宇 傅
 
Linux performance tuning & stabilization tips (mysqlconf2010)
Linux performance tuning & stabilization tips (mysqlconf2010)Linux performance tuning & stabilization tips (mysqlconf2010)
Linux performance tuning & stabilization tips (mysqlconf2010)
Yoshinori Matsunobu
 

What's hot (20)

ProdSec: A Technical Approach
ProdSec: A Technical ApproachProdSec: A Technical Approach
ProdSec: A Technical Approach
 
Datadog- Monitoring In Motion
Datadog- Monitoring In Motion Datadog- Monitoring In Motion
Datadog- Monitoring In Motion
 
Distributed Counters in Cassandra (Cassandra Summit 2010)
Distributed Counters in Cassandra (Cassandra Summit 2010)Distributed Counters in Cassandra (Cassandra Summit 2010)
Distributed Counters in Cassandra (Cassandra Summit 2010)
 
Fun with the Hak5 Rubber Ducky
Fun with the Hak5 Rubber DuckyFun with the Hak5 Rubber Ducky
Fun with the Hak5 Rubber Ducky
 
Sticky Keys to the Kingdom
Sticky Keys to the KingdomSticky Keys to the Kingdom
Sticky Keys to the Kingdom
 
A Brief History of Cryptographic Failures
A Brief History of Cryptographic FailuresA Brief History of Cryptographic Failures
A Brief History of Cryptographic Failures
 
MongoDB World 2019: MongoDB Read Isolation: Making Your Reads Clean, Committe...
MongoDB World 2019: MongoDB Read Isolation: Making Your Reads Clean, Committe...MongoDB World 2019: MongoDB Read Isolation: Making Your Reads Clean, Committe...
MongoDB World 2019: MongoDB Read Isolation: Making Your Reads Clean, Committe...
 
NameNode Analytics - Querying HDFS Namespace in Real Time
NameNode Analytics - Querying HDFS Namespace in Real TimeNameNode Analytics - Querying HDFS Namespace in Real Time
NameNode Analytics - Querying HDFS Namespace in Real Time
 
CNIT 121: 17 Remediation Introduction (Part 1)
CNIT 121: 17 Remediation Introduction (Part 1)CNIT 121: 17 Remediation Introduction (Part 1)
CNIT 121: 17 Remediation Introduction (Part 1)
 
No Easy Breach DerbyCon 2016
No Easy Breach DerbyCon 2016No Easy Breach DerbyCon 2016
No Easy Breach DerbyCon 2016
 
EDB Postgres DBA Best Practices
EDB Postgres DBA Best PracticesEDB Postgres DBA Best Practices
EDB Postgres DBA Best Practices
 
WAF Bypass Techniques - Using HTTP Standard and Web Servers’ Behaviour
WAF Bypass Techniques - Using HTTP Standard and Web Servers’ BehaviourWAF Bypass Techniques - Using HTTP Standard and Web Servers’ Behaviour
WAF Bypass Techniques - Using HTTP Standard and Web Servers’ Behaviour
 
Domino Server Health - Monitoring and Managing
 Domino Server Health - Monitoring and Managing Domino Server Health - Monitoring and Managing
Domino Server Health - Monitoring and Managing
 
C* Summit 2013: The World's Next Top Data Model by Patrick McFadin
C* Summit 2013: The World's Next Top Data Model by Patrick McFadinC* Summit 2013: The World's Next Top Data Model by Patrick McFadin
C* Summit 2013: The World's Next Top Data Model by Patrick McFadin
 
lotus domino questions answers 2017
lotus domino questions answers 2017lotus domino questions answers 2017
lotus domino questions answers 2017
 
Architecture Sustaining LINE Sticker services
Architecture Sustaining LINE Sticker servicesArchitecture Sustaining LINE Sticker services
Architecture Sustaining LINE Sticker services
 
Extending Flink State Serialization for Better Performance and Smaller Checkp...
Extending Flink State Serialization for Better Performance and Smaller Checkp...Extending Flink State Serialization for Better Performance and Smaller Checkp...
Extending Flink State Serialization for Better Performance and Smaller Checkp...
 
HTTP Request Smuggling via higher HTTP versions
HTTP Request Smuggling via higher HTTP versionsHTTP Request Smuggling via higher HTTP versions
HTTP Request Smuggling via higher HTTP versions
 
The Volcano/Cascades Optimizer
The Volcano/Cascades OptimizerThe Volcano/Cascades Optimizer
The Volcano/Cascades Optimizer
 
Linux performance tuning & stabilization tips (mysqlconf2010)
Linux performance tuning & stabilization tips (mysqlconf2010)Linux performance tuning & stabilization tips (mysqlconf2010)
Linux performance tuning & stabilization tips (mysqlconf2010)
 

Similar to Eat my data

Edubooktraining
EdubooktrainingEdubooktraining
Edubooktraining
norhloudspeaker
 
Java File I/O Performance Analysis - Part I - JCConf 2018
Java File I/O Performance Analysis - Part I - JCConf 2018Java File I/O Performance Analysis - Part I - JCConf 2018
Java File I/O Performance Analysis - Part I - JCConf 2018
Michael Fong
 
Sequential file programming patterns and performance with .net
Sequential  file programming patterns and performance with .netSequential  file programming patterns and performance with .net
Sequential file programming patterns and performance with .net
Michael Pavlovsky
 
Latihan8 comp-forensic-bab5
Latihan8 comp-forensic-bab5Latihan8 comp-forensic-bab5
Latihan8 comp-forensic-bab5sabtolinux
 
High Availability in 37 Easy Steps
High Availability in 37 Easy StepsHigh Availability in 37 Easy Steps
High Availability in 37 Easy Steps
Tim Serong
 
The care and feeding of a MySQL database
The care and feeding of a MySQL databaseThe care and feeding of a MySQL database
The care and feeding of a MySQL database
Dave Stokes
 
Purdue CS354 Operating Systems 2008
Purdue CS354 Operating Systems 2008Purdue CS354 Operating Systems 2008
Purdue CS354 Operating Systems 2008
guestd9065
 
file_c.pdf
file_c.pdffile_c.pdf
file_c.pdf
Osmania University
 
Filehandlinging cp2
Filehandlinging cp2Filehandlinging cp2
Filehandlinging cp2
Tanmay Baranwal
 
Keeping data-safe-webinar-2010-11-01
Keeping data-safe-webinar-2010-11-01Keeping data-safe-webinar-2010-11-01
Keeping data-safe-webinar-2010-11-01MongoDB
 
Troubleshooting: The Two Laws - IXIASOFT User Conference 2016
Troubleshooting: The Two Laws - IXIASOFT User Conference 2016Troubleshooting: The Two Laws - IXIASOFT User Conference 2016
Troubleshooting: The Two Laws - IXIASOFT User Conference 2016
IXIASOFT
 
File Handling In C++(OOPs))
File Handling In C++(OOPs))File Handling In C++(OOPs))
File Handling In C++(OOPs))
Papu Kumar
 
Mastering InnoDB Diagnostics
Mastering InnoDB DiagnosticsMastering InnoDB Diagnostics
Mastering InnoDB Diagnostics
guest8212a5
 
Harrison fisk masteringinnodb-diagnostics
Harrison fisk masteringinnodb-diagnosticsHarrison fisk masteringinnodb-diagnostics
Harrison fisk masteringinnodb-diagnosticsguest8212a5
 
Computer basics--basic comp-oper
Computer basics--basic comp-operComputer basics--basic comp-oper
Computer basics--basic comp-oper
Sabbir Alam
 
Ungooglable
UngooglableUngooglable
Ungooglable
Elizabeth Leddy
 

Similar to Eat my data (20)

Edubooktraining
EdubooktrainingEdubooktraining
Edubooktraining
 
Java File I/O Performance Analysis - Part I - JCConf 2018
Java File I/O Performance Analysis - Part I - JCConf 2018Java File I/O Performance Analysis - Part I - JCConf 2018
Java File I/O Performance Analysis - Part I - JCConf 2018
 
Sequential file programming patterns and performance with .net
Sequential  file programming patterns and performance with .netSequential  file programming patterns and performance with .net
Sequential file programming patterns and performance with .net
 
Latihan8 comp-forensic-bab5
Latihan8 comp-forensic-bab5Latihan8 comp-forensic-bab5
Latihan8 comp-forensic-bab5
 
High Availability in 37 Easy Steps
High Availability in 37 Easy StepsHigh Availability in 37 Easy Steps
High Availability in 37 Easy Steps
 
The care and feeding of a MySQL database
The care and feeding of a MySQL databaseThe care and feeding of a MySQL database
The care and feeding of a MySQL database
 
Purdue CS354 Operating Systems 2008
Purdue CS354 Operating Systems 2008Purdue CS354 Operating Systems 2008
Purdue CS354 Operating Systems 2008
 
Measuring Firebird Disk I/O
Measuring Firebird Disk I/OMeasuring Firebird Disk I/O
Measuring Firebird Disk I/O
 
Ch23 system administration
Ch23 system administration Ch23 system administration
Ch23 system administration
 
file_c.pdf
file_c.pdffile_c.pdf
file_c.pdf
 
Filehandlinging cp2
Filehandlinging cp2Filehandlinging cp2
Filehandlinging cp2
 
Keeping data-safe-webinar-2010-11-01
Keeping data-safe-webinar-2010-11-01Keeping data-safe-webinar-2010-11-01
Keeping data-safe-webinar-2010-11-01
 
File Handling In C++
File Handling In C++File Handling In C++
File Handling In C++
 
Ext 0523
Ext 0523Ext 0523
Ext 0523
 
Troubleshooting: The Two Laws - IXIASOFT User Conference 2016
Troubleshooting: The Two Laws - IXIASOFT User Conference 2016Troubleshooting: The Two Laws - IXIASOFT User Conference 2016
Troubleshooting: The Two Laws - IXIASOFT User Conference 2016
 
File Handling In C++(OOPs))
File Handling In C++(OOPs))File Handling In C++(OOPs))
File Handling In C++(OOPs))
 
Mastering InnoDB Diagnostics
Mastering InnoDB DiagnosticsMastering InnoDB Diagnostics
Mastering InnoDB Diagnostics
 
Harrison fisk masteringinnodb-diagnostics
Harrison fisk masteringinnodb-diagnosticsHarrison fisk masteringinnodb-diagnostics
Harrison fisk masteringinnodb-diagnostics
 
Computer basics--basic comp-oper
Computer basics--basic comp-operComputer basics--basic comp-oper
Computer basics--basic comp-oper
 
Ungooglable
UngooglableUngooglable
Ungooglable
 

Recently uploaded

AI for Every Business: Unlocking Your Product's Universal Potential by VP of ...
AI for Every Business: Unlocking Your Product's Universal Potential by VP of ...AI for Every Business: Unlocking Your Product's Universal Potential by VP of ...
AI for Every Business: Unlocking Your Product's Universal Potential by VP of ...
Product School
 
Builder.ai Founder Sachin Dev Duggal's Strategic Approach to Create an Innova...
Builder.ai Founder Sachin Dev Duggal's Strategic Approach to Create an Innova...Builder.ai Founder Sachin Dev Duggal's Strategic Approach to Create an Innova...
Builder.ai Founder Sachin Dev Duggal's Strategic Approach to Create an Innova...
Ramesh Iyer
 
ODC, Data Fabric and Architecture User Group
ODC, Data Fabric and Architecture User GroupODC, Data Fabric and Architecture User Group
ODC, Data Fabric and Architecture User Group
CatarinaPereira64715
 
Essentials of Automations: Optimizing FME Workflows with Parameters
Essentials of Automations: Optimizing FME Workflows with ParametersEssentials of Automations: Optimizing FME Workflows with Parameters
Essentials of Automations: Optimizing FME Workflows with Parameters
Safe Software
 
Neuro-symbolic is not enough, we need neuro-*semantic*
Neuro-symbolic is not enough, we need neuro-*semantic*Neuro-symbolic is not enough, we need neuro-*semantic*
Neuro-symbolic is not enough, we need neuro-*semantic*
Frank van Harmelen
 
JMeter webinar - integration with InfluxDB and Grafana
JMeter webinar - integration with InfluxDB and GrafanaJMeter webinar - integration with InfluxDB and Grafana
JMeter webinar - integration with InfluxDB and Grafana
RTTS
 
Smart TV Buyer Insights Survey 2024 by 91mobiles.pdf
Smart TV Buyer Insights Survey 2024 by 91mobiles.pdfSmart TV Buyer Insights Survey 2024 by 91mobiles.pdf
Smart TV Buyer Insights Survey 2024 by 91mobiles.pdf
91mobiles
 
To Graph or Not to Graph Knowledge Graph Architectures and LLMs
To Graph or Not to Graph Knowledge Graph Architectures and LLMsTo Graph or Not to Graph Knowledge Graph Architectures and LLMs
To Graph or Not to Graph Knowledge Graph Architectures and LLMs
Paul Groth
 
State of ICS and IoT Cyber Threat Landscape Report 2024 preview
State of ICS and IoT Cyber Threat Landscape Report 2024 previewState of ICS and IoT Cyber Threat Landscape Report 2024 preview
State of ICS and IoT Cyber Threat Landscape Report 2024 preview
Prayukth K V
 
Connector Corner: Automate dynamic content and events by pushing a button
Connector Corner: Automate dynamic content and events by pushing a buttonConnector Corner: Automate dynamic content and events by pushing a button
Connector Corner: Automate dynamic content and events by pushing a button
DianaGray10
 
Knowledge engineering: from people to machines and back
Knowledge engineering: from people to machines and backKnowledge engineering: from people to machines and back
Knowledge engineering: from people to machines and back
Elena Simperl
 
Leading Change strategies and insights for effective change management pdf 1.pdf
Leading Change strategies and insights for effective change management pdf 1.pdfLeading Change strategies and insights for effective change management pdf 1.pdf
Leading Change strategies and insights for effective change management pdf 1.pdf
OnBoard
 
UiPath Test Automation using UiPath Test Suite series, part 4
UiPath Test Automation using UiPath Test Suite series, part 4UiPath Test Automation using UiPath Test Suite series, part 4
UiPath Test Automation using UiPath Test Suite series, part 4
DianaGray10
 
GDG Cloud Southlake #33: Boule & Rebala: Effective AppSec in SDLC using Deplo...
GDG Cloud Southlake #33: Boule & Rebala: Effective AppSec in SDLC using Deplo...GDG Cloud Southlake #33: Boule & Rebala: Effective AppSec in SDLC using Deplo...
GDG Cloud Southlake #33: Boule & Rebala: Effective AppSec in SDLC using Deplo...
James Anderson
 
Epistemic Interaction - tuning interfaces to provide information for AI support
Epistemic Interaction - tuning interfaces to provide information for AI supportEpistemic Interaction - tuning interfaces to provide information for AI support
Epistemic Interaction - tuning interfaces to provide information for AI support
Alan Dix
 
Mission to Decommission: Importance of Decommissioning Products to Increase E...
Mission to Decommission: Importance of Decommissioning Products to Increase E...Mission to Decommission: Importance of Decommissioning Products to Increase E...
Mission to Decommission: Importance of Decommissioning Products to Increase E...
Product School
 
Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...
Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...
Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...
Jeffrey Haguewood
 
Accelerate your Kubernetes clusters with Varnish Caching
Accelerate your Kubernetes clusters with Varnish CachingAccelerate your Kubernetes clusters with Varnish Caching
Accelerate your Kubernetes clusters with Varnish Caching
Thijs Feryn
 
Assuring Contact Center Experiences for Your Customers With ThousandEyes
Assuring Contact Center Experiences for Your Customers With ThousandEyesAssuring Contact Center Experiences for Your Customers With ThousandEyes
Assuring Contact Center Experiences for Your Customers With ThousandEyes
ThousandEyes
 
FIDO Alliance Osaka Seminar: Passkeys and the Road Ahead.pdf
FIDO Alliance Osaka Seminar: Passkeys and the Road Ahead.pdfFIDO Alliance Osaka Seminar: Passkeys and the Road Ahead.pdf
FIDO Alliance Osaka Seminar: Passkeys and the Road Ahead.pdf
FIDO Alliance
 

Recently uploaded (20)

AI for Every Business: Unlocking Your Product's Universal Potential by VP of ...
AI for Every Business: Unlocking Your Product's Universal Potential by VP of ...AI for Every Business: Unlocking Your Product's Universal Potential by VP of ...
AI for Every Business: Unlocking Your Product's Universal Potential by VP of ...
 
Builder.ai Founder Sachin Dev Duggal's Strategic Approach to Create an Innova...
Builder.ai Founder Sachin Dev Duggal's Strategic Approach to Create an Innova...Builder.ai Founder Sachin Dev Duggal's Strategic Approach to Create an Innova...
Builder.ai Founder Sachin Dev Duggal's Strategic Approach to Create an Innova...
 
ODC, Data Fabric and Architecture User Group
ODC, Data Fabric and Architecture User GroupODC, Data Fabric and Architecture User Group
ODC, Data Fabric and Architecture User Group
 
Essentials of Automations: Optimizing FME Workflows with Parameters
Essentials of Automations: Optimizing FME Workflows with ParametersEssentials of Automations: Optimizing FME Workflows with Parameters
Essentials of Automations: Optimizing FME Workflows with Parameters
 
Neuro-symbolic is not enough, we need neuro-*semantic*
Neuro-symbolic is not enough, we need neuro-*semantic*Neuro-symbolic is not enough, we need neuro-*semantic*
Neuro-symbolic is not enough, we need neuro-*semantic*
 
JMeter webinar - integration with InfluxDB and Grafana
JMeter webinar - integration with InfluxDB and GrafanaJMeter webinar - integration with InfluxDB and Grafana
JMeter webinar - integration with InfluxDB and Grafana
 
Smart TV Buyer Insights Survey 2024 by 91mobiles.pdf
Smart TV Buyer Insights Survey 2024 by 91mobiles.pdfSmart TV Buyer Insights Survey 2024 by 91mobiles.pdf
Smart TV Buyer Insights Survey 2024 by 91mobiles.pdf
 
To Graph or Not to Graph Knowledge Graph Architectures and LLMs
To Graph or Not to Graph Knowledge Graph Architectures and LLMsTo Graph or Not to Graph Knowledge Graph Architectures and LLMs
To Graph or Not to Graph Knowledge Graph Architectures and LLMs
 
State of ICS and IoT Cyber Threat Landscape Report 2024 preview
State of ICS and IoT Cyber Threat Landscape Report 2024 previewState of ICS and IoT Cyber Threat Landscape Report 2024 preview
State of ICS and IoT Cyber Threat Landscape Report 2024 preview
 
Connector Corner: Automate dynamic content and events by pushing a button
Connector Corner: Automate dynamic content and events by pushing a buttonConnector Corner: Automate dynamic content and events by pushing a button
Connector Corner: Automate dynamic content and events by pushing a button
 
Knowledge engineering: from people to machines and back
Knowledge engineering: from people to machines and backKnowledge engineering: from people to machines and back
Knowledge engineering: from people to machines and back
 
Leading Change strategies and insights for effective change management pdf 1.pdf
Leading Change strategies and insights for effective change management pdf 1.pdfLeading Change strategies and insights for effective change management pdf 1.pdf
Leading Change strategies and insights for effective change management pdf 1.pdf
 
UiPath Test Automation using UiPath Test Suite series, part 4
UiPath Test Automation using UiPath Test Suite series, part 4UiPath Test Automation using UiPath Test Suite series, part 4
UiPath Test Automation using UiPath Test Suite series, part 4
 
GDG Cloud Southlake #33: Boule & Rebala: Effective AppSec in SDLC using Deplo...
GDG Cloud Southlake #33: Boule & Rebala: Effective AppSec in SDLC using Deplo...GDG Cloud Southlake #33: Boule & Rebala: Effective AppSec in SDLC using Deplo...
GDG Cloud Southlake #33: Boule & Rebala: Effective AppSec in SDLC using Deplo...
 
Epistemic Interaction - tuning interfaces to provide information for AI support
Epistemic Interaction - tuning interfaces to provide information for AI supportEpistemic Interaction - tuning interfaces to provide information for AI support
Epistemic Interaction - tuning interfaces to provide information for AI support
 
Mission to Decommission: Importance of Decommissioning Products to Increase E...
Mission to Decommission: Importance of Decommissioning Products to Increase E...Mission to Decommission: Importance of Decommissioning Products to Increase E...
Mission to Decommission: Importance of Decommissioning Products to Increase E...
 
Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...
Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...
Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...
 
Accelerate your Kubernetes clusters with Varnish Caching
Accelerate your Kubernetes clusters with Varnish CachingAccelerate your Kubernetes clusters with Varnish Caching
Accelerate your Kubernetes clusters with Varnish Caching
 
Assuring Contact Center Experiences for Your Customers With ThousandEyes
Assuring Contact Center Experiences for Your Customers With ThousandEyesAssuring Contact Center Experiences for Your Customers With ThousandEyes
Assuring Contact Center Experiences for Your Customers With ThousandEyes
 
FIDO Alliance Osaka Seminar: Passkeys and the Road Ahead.pdf
FIDO Alliance Osaka Seminar: Passkeys and the Road Ahead.pdfFIDO Alliance Osaka Seminar: Passkeys and the Road Ahead.pdf
FIDO Alliance Osaka Seminar: Passkeys and the Road Ahead.pdf
 

Eat my data

  • 1. Eat My Data: (now with 20% more rant!) How everybody gets file I/O wrong Stewart Smith [email_address] Senior Software Engineer, MySQL Cluster MySQL AB
  • 2.
  • 3.
  • 4.
  • 5.
  • 6.
  • 7.
  • 8.
  • 9.
  • 10.
  • 11. A world without failure
  • 12.
  • 13.
  • 14.
  • 15.
  • 16.
  • 17.
  • 18.
  • 19.
  • 20.
  • 21.
  • 22.
  • 24.
  • 25.
  • 26.
  • 27.
  • 28.
  • 29.
  • 30.
  • 31.
  • 32.
  • 33.
  • 34.
  • 35.
  • 36.
  • 37.
  • 38.
  • 39.
  • 40.
  • 41.
  • 42.
  • 43.
  • 44.
  • 45.
  • 46.
  • 47.
  • 48.
  • 49.
  • 50.
  • 51.
  • 52.
  • 53.
  • 54.
  • 55.
  • 56.
  • 57.
  • 58.
  • 59.
  • 60.
  • 61.
  • 62.
  • 64.
  • 65.
  • 66.
  • 67.
  • 68.
  • 69.
  • 70.
  • 71.
  • 72.
  • 73.
  • 74.
  • 75.
  • 76. Now all is good with the world...
  • 77.
  • 78.
  • 79.
  • 80.
  • 81.
  • 82.
  • 83.
  • 84.
  • 85.
  • 86.
  • 87.
  • 88.
  • 89.
  • 90.
  • 91.
  • 92.
  • 93.
  • 94. gint common_save_xml(xmlDocPtr doc, gchar *filename) { FILE *fp; char *xmlbuf; int fd, n; fp = g_fopen(filename, &quot;w&quot;); if(NULL == fp) return -1; xmlDocDumpFormatMemory(doc, (xmlChar **)&xmlbuf, &n, TRUE); if(n <= 0) { errno = ENOMEM; return -1; } if(fwrite(xmlbuf, sizeof (xmlChar), n, fp) < n) { xmlFree (xmlbuf); return -1; } xmlFree (xmlbuf); /* flush user-space buffers */ if (fflush (fp) != 0) return -1; if ((fd = fileno (fp)) == -1) return -1; #ifdef HAVE_FSYNC /* sync kernel-space buffers to disk */ if (fsync (fd) == -1) return -1; #endif fclose(fp); return 0; }
  • 95.
  • 96.
  • 97.
  • 98.
  • 99.
  • 100.
  • 101.
  • 102.
  • 103.
  • 104.
  • 105.
  • 106.
  • 107.
  • 108.
  • 109.
  • 110.
  • 111.
  • 112.
  • 113.
  • 114.
  • 115.
  • 116.
  • 117. #ifdef HAVE_DARWIN_THREADS # ifdef F_FULLFSYNC /* This executable has been compiled on Mac OS X 10.3 or later. Assume that F_FULLFSYNC is available at run-time. */ srv_have_fullfsync = TRUE; # else /* F_FULLFSYNC */ /* This executable has been compiled on Mac OS X 10.2 or earlier. Determine if the executable is running on Mac OS X 10.3 or later. */ struct utsname utsname; if (uname(&utsname)) { fputs(&quot;InnoDB: cannot determine Mac OS X version!&quot;, stderr); } else { srv_have_fullfsync = strcmp(utsname.release, &quot;7.&quot;) >= 0; } if (!srv_have_fullfsync) { fputs(&quot;InnoDB: On Mac OS X, fsync() may be&quot; &quot; broken on internal drives,&quot; &quot;InnoDB: making transactions unsafe!&quot;, stderr); } # endif /* F_FULLFSYNC */ #endif /* HAVE_DARWIN_THREADS */
  • 118. #if defined(HAVE_DARWIN_THREADS) # ifndef F_FULLFSYNC /* The following definition is from the Mac OS X 10.3 <sys/fcntl.h> */ # define F_FULLFSYNC 51 /* fsync + ask the drive to flush to the media */ # elif F_FULLFSYNC != 51 # error &quot;F_FULLFSYNC != 51: ABI incompatibility with Mac OS X 10.3&quot; # endif /* Apple has disabled fsync() for internal disk drives in OS X. That caused corruption for a user when he tested a power outage. Let us in OS X use a nonstandard flush method recommended by an Apple engineer. */ if (!srv_have_fullfsync) { /* If we are not on an operating system that supports this, then fall back to a plain fsync. */ ret = fsync(file); } else { ret = fcntl(file, F_FULLFSYNC, NULL); if (ret) { /* If we are not on a file system that supports this, then fall back to a plain fsync. */ ret = fsync(file); } } #elif HAVE_FDATASYNC ret = fdatasync(file); #else /* fprintf(stderr, &quot;Flushing to file %p&quot;, file); */ ret = fsync(file); #endif
  • 119.
  • 121.
  • 122.
  • 123.
  • 124.
  • 125.
  • 126.
  • 127.
  • 128.
  • 129.
  • 130.
  • 131.
  • 132.
  • 133.
  • 134.
  • 135.
  • 136.
  • 137.
  • 138.
  • 139.
  • 140.
  • 141.
  • 142.
  • 143.
  • 144.
  • 145.
  • 146.
  • 147.
  • 148.
  • 149. inode Infos Direct blocks Indirect Blocks Double indirect blocks
  • 150.
  • 151.
  • 152.
  • 153.
  • 154.
  • 155.
  • 156.
  • 157.
  • 158.
  • 159.
  • 160.
  • 161.
  • 162.
  • 163.
  • 164.
  • 165.
  • 166.
  • 168.