Practical information on how to Optimize Virtual Machines for High Performance by Boyan Krosnov, Chief Product Officer at StorPool Storage
Presentation delivered at OpenNebula TechDay Sofia on 25-th of February 2016
Linux Memory Basics for SysAdmins - ChinaNetCloud TrainingChinaNetCloud
ChinaNetCloud training for Linux Memory Basics for SysAdmins.
This is an introduction to general Linux memory for troubleshooting, monitoring, and basic understanding.
Practical information on how to Optimize Virtual Machines for High Performance by Boyan Krosnov, Chief Product Officer at StorPool Storage
Presentation delivered at OpenNebula TechDay Sofia on 25-th of February 2016
Linux Memory Basics for SysAdmins - ChinaNetCloud TrainingChinaNetCloud
ChinaNetCloud training for Linux Memory Basics for SysAdmins.
This is an introduction to general Linux memory for troubleshooting, monitoring, and basic understanding.
Talk given on state of NUMA with Java databases such as Cassandra and how it can improved / ameliorated, and compared with traditional storage engines.
This talk is about a new interface to get information about processes, called task_diag, which we developed.
Currently /proc file system is used to get information about the processes running on the system. All information are presented as text files, which is convenient for humans, but not for programs such as ps and top. This incurs significant delays, especially on a systems with lots of containers running, which is frequently the case nowdays.
Ideally, tools such top and ps would get information in binary format, and use flexible means to specify which kinds of information and for which tasks is required. Presented is a new interface with all these features, called task_diag.
task_diag is based on netlink sockets and looks like socket-diag, which is used to get information about sockets. It uses the request-response model. An request specifies a set of processes and required properties for them. A response contains requested information and can be divided into a few netlink packets if it's too long.
The task diag is much faster than the /proc file system. For example, when reading from /proc, ps opens, reads, and closes many files -- and iterates this for every single processes. With task_diag, it's just sending a request and getting a response.
Except for ps and top, the proposed interface is to be used by CRIU, a containers checkpoint/restore and live migration mechanism. Also, developers of perf tool found that it can be useful to them and implemented a prototype which show a big performance improvements in case of using task_diag instead of procfs.
Our performance measurements show that the ps tool works at least four times faster if task_diag is used instead of procfs.
How Can OpenNebula Fit Your Needs: A European Project FeedbackNETWAYS
BonFIRE is an european project which aims at providing a ”multi-site cloud facility for applications, services and systems research and experimentation”. Grouping different research cloud providers behind a common set of tools, APIs and services, it enables users to run their experiment against a heterogeneous set of infrastructure, hypervisors, networks, etc …
BonFIRE, and thus the (OpenNebula) testbeds, provide a relatively small set of images used to boot VMs. However, the experimental nature of BonFIRE projects results in a big ”turnover” of running VMs. Lot of VMs are used for a time period between a few hours and a few days, and an experiment startup can trigger deployment of many VMs at same time on a small set of OpenNebula workers, which does not correspond to usual Cloud workflow.
Default OpenNebula is not optimized for such usecase (small amount of worker nodes, high VMs turnover). However, thanks to its ability to be easily modified at each level of a Cloud deployment workflow, OpenNebula has been tuned to make it fit better with BonFIRE deployment process. This presentation will explain how to change OpenNebula TM and VMM to improve the parrallel deployment of many VMs in a short amount of time, reducing time needed to deploy an experiment to its lowest without lot of expensive hardware.
OpenNebulaConf2015 2.14 Cloud Service Experience in TeideHPC Infrastructure -...OpenNebula Project
TeideHPC is a High Performance Computing infrastructure, used for research and development tasks in a variety of areas like weather forecasting, astrophysics, CFD or bioinformatics. We are also involved in other fields of study, less related with R&D such as render or streaming services, and other with changing requirements, such the evaluation of pilot environments for open government/open data, research institutes or companies from the engineering sector.
In those jobs we have found OpenNebula as a great solution with many capabilities but also with some limitations. The aim of these session is to show our experience with OpenNebula, from a technical viewpoint, presenting some topics on deployment/configuration with Cobbler/Chef, Infiniband virtualization with SR-IOV, Power management and Cloud bursting limitations.
Author Biography
Carlos Ignacio González Vila is a Software and System Engineer. Worked in the Research Support Computing Service in the same University for three years, doing a wide variety of tasks such as development of research applications for data analysis, project management or system administration of the clusters and supercomputers of the University.
Since 2012 he is working as a High Performance Computing System Administrator in the Technological and Renewable Energies Institute (ITER) of Tenerife (Canary Islands), involved in the TeideHPC supercomputer proyect, a computing infrastructure with more than 1100 computing nodes and nearly 40 TB of RAM memory. His job includes storage management, ethernet and infiniband network, scientific application and cloud infrastructure.
OpenNebulaConf 2016 - The DRBD SDS for OpenNebula by Philipp Reisner, LINBITOpenNebula Project
You will learn what DRBD is, where it came from in its 15 years of existence. How it evolved into a software defined storage solution interesting for users of OpenNebula and why it is very well suited for hyperconverged deployment architectures. The presentation will contain IO performance results and (if time permits) a live demo.
How can OpenNebula fit your needs - OpenNebulaConf 2013 Maxence Dunnewind
In the scope of a European Project (BonFIRE - www.bonfire-project.eu ), I had to tune openNebula to fit our requirement that are unusual in a private cloud environment (small hardware, small number of base images, but lot of vms created).
These slides explain how, thanks to how OpenNebula enables administrators to tune it, I updated the transfer manager scripts to improve our deployment speed by almost 8.
MariaDB/MySQL pitfalls - And how to come out again...FromDual GmbH
During the last conferences the audience was asking for more war stories than just new features.
In this presentation we have a look at the most often seen problems as a MariaDB/MySQL consultant in field.
Talk given on state of NUMA with Java databases such as Cassandra and how it can improved / ameliorated, and compared with traditional storage engines.
This talk is about a new interface to get information about processes, called task_diag, which we developed.
Currently /proc file system is used to get information about the processes running on the system. All information are presented as text files, which is convenient for humans, but not for programs such as ps and top. This incurs significant delays, especially on a systems with lots of containers running, which is frequently the case nowdays.
Ideally, tools such top and ps would get information in binary format, and use flexible means to specify which kinds of information and for which tasks is required. Presented is a new interface with all these features, called task_diag.
task_diag is based on netlink sockets and looks like socket-diag, which is used to get information about sockets. It uses the request-response model. An request specifies a set of processes and required properties for them. A response contains requested information and can be divided into a few netlink packets if it's too long.
The task diag is much faster than the /proc file system. For example, when reading from /proc, ps opens, reads, and closes many files -- and iterates this for every single processes. With task_diag, it's just sending a request and getting a response.
Except for ps and top, the proposed interface is to be used by CRIU, a containers checkpoint/restore and live migration mechanism. Also, developers of perf tool found that it can be useful to them and implemented a prototype which show a big performance improvements in case of using task_diag instead of procfs.
Our performance measurements show that the ps tool works at least four times faster if task_diag is used instead of procfs.
How Can OpenNebula Fit Your Needs: A European Project FeedbackNETWAYS
BonFIRE is an european project which aims at providing a ”multi-site cloud facility for applications, services and systems research and experimentation”. Grouping different research cloud providers behind a common set of tools, APIs and services, it enables users to run their experiment against a heterogeneous set of infrastructure, hypervisors, networks, etc …
BonFIRE, and thus the (OpenNebula) testbeds, provide a relatively small set of images used to boot VMs. However, the experimental nature of BonFIRE projects results in a big ”turnover” of running VMs. Lot of VMs are used for a time period between a few hours and a few days, and an experiment startup can trigger deployment of many VMs at same time on a small set of OpenNebula workers, which does not correspond to usual Cloud workflow.
Default OpenNebula is not optimized for such usecase (small amount of worker nodes, high VMs turnover). However, thanks to its ability to be easily modified at each level of a Cloud deployment workflow, OpenNebula has been tuned to make it fit better with BonFIRE deployment process. This presentation will explain how to change OpenNebula TM and VMM to improve the parrallel deployment of many VMs in a short amount of time, reducing time needed to deploy an experiment to its lowest without lot of expensive hardware.
OpenNebulaConf2015 2.14 Cloud Service Experience in TeideHPC Infrastructure -...OpenNebula Project
TeideHPC is a High Performance Computing infrastructure, used for research and development tasks in a variety of areas like weather forecasting, astrophysics, CFD or bioinformatics. We are also involved in other fields of study, less related with R&D such as render or streaming services, and other with changing requirements, such the evaluation of pilot environments for open government/open data, research institutes or companies from the engineering sector.
In those jobs we have found OpenNebula as a great solution with many capabilities but also with some limitations. The aim of these session is to show our experience with OpenNebula, from a technical viewpoint, presenting some topics on deployment/configuration with Cobbler/Chef, Infiniband virtualization with SR-IOV, Power management and Cloud bursting limitations.
Author Biography
Carlos Ignacio González Vila is a Software and System Engineer. Worked in the Research Support Computing Service in the same University for three years, doing a wide variety of tasks such as development of research applications for data analysis, project management or system administration of the clusters and supercomputers of the University.
Since 2012 he is working as a High Performance Computing System Administrator in the Technological and Renewable Energies Institute (ITER) of Tenerife (Canary Islands), involved in the TeideHPC supercomputer proyect, a computing infrastructure with more than 1100 computing nodes and nearly 40 TB of RAM memory. His job includes storage management, ethernet and infiniband network, scientific application and cloud infrastructure.
OpenNebulaConf 2016 - The DRBD SDS for OpenNebula by Philipp Reisner, LINBITOpenNebula Project
You will learn what DRBD is, where it came from in its 15 years of existence. How it evolved into a software defined storage solution interesting for users of OpenNebula and why it is very well suited for hyperconverged deployment architectures. The presentation will contain IO performance results and (if time permits) a live demo.
How can OpenNebula fit your needs - OpenNebulaConf 2013 Maxence Dunnewind
In the scope of a European Project (BonFIRE - www.bonfire-project.eu ), I had to tune openNebula to fit our requirement that are unusual in a private cloud environment (small hardware, small number of base images, but lot of vms created).
These slides explain how, thanks to how OpenNebula enables administrators to tune it, I updated the transfer manager scripts to improve our deployment speed by almost 8.
MariaDB/MySQL pitfalls - And how to come out again...FromDual GmbH
During the last conferences the audience was asking for more war stories than just new features.
In this presentation we have a look at the most often seen problems as a MariaDB/MySQL consultant in field.
MariaDB / MySQL tripping hazard and how to get out again?FromDual GmbH
MySQL is NOT MariaDB, KISS, MyISAM Table Locking, big ibdata1, Noisy Neighbours, No BLOBs in a RDBMS, Disk full!, Locking, Crash, Out of Memory, Oom, Restart, Crash with Stacktrace, Tables without a Primary Key, Server has gone away! NUMA, Slave Lag, Sending data, Too many connections, Simple Query Tuning, Galera Cluster, Progress bar
A presentation about how to make MySQL highly available, presented at the San Francisco MySQL Meetup (http://www.sfmysql.org/events/15760472/) on January 26th, 2011.
A video recording of this presentation is available from Ustream: http://ustre.am/fyLk
This tutorial covers all parallel replication implementation in MariaDB 10.0 and 10.1 and MySQL 5.6, 5.7 and 8.0 (including how it works in Group Replication).
MySQL and MariaDB have different types of parallel replication. In this tutorial, we present the different implementations that allow us to understand their limitations and tuning parameters. We cover how to make parallel replication faster and what to avoid for maximizing its benefits. We also present tests from Booking.com workloads.
Some of the subjects that are covered are group commit and optimistic parallel replication in MariaDB, the parallelism interval of MySQL and its Write Set optimization, and the ?slowing down the master to speed up the slave? optimization.
After this tutorial, you will know everything you need to implement and tune parallel replication in your environment. But more importantly, we will show how you can test parallel replication benefit in a non-disruptive way before deployment.
Scaling up and accelerating Drupal 8 with NoSQLOSInet
Drupal 8 can scale well and serve pages fast to many users, especially by offloading parts of the work load from the main SQL database to NoSQL solutions.
This presentation describes the strategies and technologies usable to achieve such gains, including specific configuration, contributed modules and custom coding strategies.
MySQL Parallel Replication: inventory, use-case and limitationsJean-François Gagné
In the last 24 months, MySQL replication speed has improved a lot thanks to implementing parallel replication. MySQL and MariaDB have different types of parallel replication; in this talk, I present in details the different implementations, with their limitations and the corresponding tuning parameters. I also present benchmark results from real Booking.com workloads. Finally, I discuss some deployments at Booking.com that benefits from parallel replication speed improvements.
This presentation describes the challenges we faced building, scaling and operating a Kubernetes cluster of more than 1000 nodes to host the Datadog applications
MySQL/MariaDB Parallel Replication: inventory, use-case and limitationsJean-François Gagné
In the last 24 months, MySQL/MariaDB replication speed has improved a lot, thanks to parallel replication. MySQL and MariaDB Server have different types of parallel replication; in this talk, I present the different implementations which will allow us to understand their limitations and tuning parameters. I am covering how to make parallel replication faster and what to avoid for maximizing its benefits. I also present benchmark results from Booking.com workloads. Finally, I discuss some deployments at Booking.com that take advantage of parallel replication speed improvements.
Coredns nodecache - A highly-available Node-cache DNS serverYann Hamon
The Kubernetes Node-local dns add-on proposes running a DNS caching server on all of a Kubernetes cluster's nodes. The suggested caching server is node-cache, a thin wrapper around CoreDNS, that handles the setup & teardown of the dummy network interface & associated IPTables rules.
Coredns-nodecache is an attempt to implement node-cache as a CoreDNS plugin, rather than a wrapper. It can be configured in a highly-available setup, and is setup using the CoreDNS Corefile
June 24, 2014. At Velocity 2014, Fastly engineer Vladimir Vuksan gave an intro to Ganglia concepts (grid, clusters, hosts) as well as an installation of a sample monitoring grid. He also goes through the following commonly used visualization tools and how they may aid in detecting issues, identifying causes, and taking corrective action:
- Cluster/Grid Views
- Aggregate graphs
- Compare Hosts
- Custom graph functionality
- Views
- Interactive graphs
- Trending
- Nagios/Alerting system integration
- How to add metrics to Ganglia
- Different export formats such as JSON, CSV, and XML
MySQL backups overview. Characteristics of every backup type, including dumps, Xtrabackup and snapshots. Planning proper backup strategies. Why and how to test backups.
MariaDB Server on macOS - FOSDEM 2022 MariaDB DevroomValeriy Kravchuk
Current MariaDB Server GA versions are formally not supported (and probably not even regularly built or tested) on macOS 10.x and 11.y. But it's relatively easy to set up the environment and build MariaDB Server from current 10.2 - 10.8 GitHub sources, with few minor issues to resolve in the process, depending on macOS and major server version used.
This talk is a summary of my related experience on 10.13 High Sierra that I had a chance to work on recently, with additional quick review of related fixed and open bugs, as well as some unique features like DTrace support that one may benefit from on macOS. Actually, studying DTrace in context of MariaDB Server troubleshooting and performance tuning was one of the goals why I started to use macOS again.
Data deduplication is a hot topic in storage and saves significant disk space for many environments, with some trade offs. We’ll discuss what deduplication is and where the Open Source solutions are versus commercial offerings. Presentation will lean towards the practical – where attendees can use it in their real world projects (what works, what doesn’t, should you use in production, etcetera).
Similar to High-availability with Galera Cluster for MySQL (20)
In this presentation we discuss the New Features of MariaDB 10.4. First we give a short overview of the MariaDB Branches and Forks. Then we talk about the announced IPO. Technically we cover topics like Authentication, Accounts, InnoDB, Optimizer improvements, Application-Time Period Tables the new Backup Stage Galera 4 and other changes...
MariaDB 10.4 became General Available (GA = ready for production) this summer. So it is time to look at the new Features in MariaDB 10.4. After a short intro about history we look for the reason of broad usage of MariaDB nowadays. Most important improvements where in User Authentication, InnoDB improvements, and Optimizer enhancements. A completely New Feature is Application-Time Period Tables. Backup got a new Locking behaviour so LVM snapshots are possible and officially supported now. And last but not least MariaDB 10.4 comes with Galera 4.
MariaDB 10.2 New Features for Developers, Administrators and DevOps. Window Functions, Common Table Expressions, Check Constraints, GeoJSON, GIS, JSON, Oracle compatibility and MariaDB Connectors
Der Datenbank-Backup ist gemacht - was nun?FromDual GmbH
* Datenbank-Backup – welcher Zweck?
* Tauglichkeit des Backup, Verifikation
* Echtdaten vollständig nutzen
* Dem Datenschutz genügen
* Material für die Entwicklung
* Automatisierung
Weltweite Produktionsdatenverwaltung mit MySQL-ReplikationFromDual GmbH
Weltweite Produktionsdatenverwaltung mit MySQL
* Ausgangslage
* Probleme die sich stellen
* Wer darf welche Daten kriegen?
* Wie werden Daten verteilt?
* Produktionsdaten zurück?
* MySQL multi-Source Replikation
* Bertriebsverantwortung
* Hochverfügbarkeit
* Sensitive Daten
* Wer darf welche Daten sehen?
* MySQL Row Filterung
* Nachträgliche Forderungen
MySQL Performance Tuning für Oracle-DBA'sFromDual GmbH
MySQL Performance Tuning
* Was ist Performance?
* Was kostet Performance?
* Tuning Massnahmen
* MySQL Konfiguration
* Wo schauen?
* Langsame Abfragen finden
* Optimiere das Query!
* Monitoring
Reading MySQL fingerprints
Every MySQL Database has its typical fingerprints. They can be seen with SHOW GLOBAL STATUS;
How you interprete them you will find in this presentation.
MySQL indexing is one of the areas where you can achieve the most performance gains. In this presentation we have a short look how we can improve MySQL performance with indexing...
Collapsing Narratives: Exploring Non-Linearity • a micro report by Rosie WellsRosie Wells
Insight: In a landscape where traditional narrative structures are giving way to fragmented and non-linear forms of storytelling, there lies immense potential for creativity and exploration.
'Collapsing Narratives: Exploring Non-Linearity' is a micro report from Rosie Wells.
Rosie Wells is an Arts & Cultural Strategist uniquely positioned at the intersection of grassroots and mainstream storytelling.
Their work is focused on developing meaningful and lasting connections that can drive social change.
Please download this presentation to enjoy the hyperlinks!
This presentation, created by Syed Faiz ul Hassan, explores the profound influence of media on public perception and behavior. It delves into the evolution of media from oral traditions to modern digital and social media platforms. Key topics include the role of media in information propagation, socialization, crisis awareness, globalization, and education. The presentation also examines media influence through agenda setting, propaganda, and manipulative techniques used by advertisers and marketers. Furthermore, it highlights the impact of surveillance enabled by media technologies on personal behavior and preferences. Through this comprehensive overview, the presentation aims to shed light on how media shapes collective consciousness and public opinion.
Mastering the Concepts Tested in the Databricks Certified Data Engineer Assoc...SkillCertProExams
• For a full set of 760+ questions. Go to
https://skillcertpro.com/product/databricks-certified-data-engineer-associate-exam-questions/
• SkillCertPro offers detailed explanations to each question which helps to understand the concepts better.
• It is recommended to score above 85% in SkillCertPro exams before attempting a real exam.
• SkillCertPro updates exam questions every 2 weeks.
• You will get life time access and life time free updates
• SkillCertPro assures 100% pass guarantee in first attempt.
Mastering the Concepts Tested in the Databricks Certified Data Engineer Assoc...
High-availability with Galera Cluster for MySQL
1. www.fromdual.com
1 / 21
High-availability with
Galera Cluster
for MySQL
FromDual Company Meeting
10. September 2014, Barcelona
by oli.sennhauser@fromdual.com
www.fromdual.com
2. www.fromdual.com
2 / 21
About FromDual GmbH
● FromDual provides neutral and independent:
● Consulting for MySQL, Galera Cluster, MariaDB and
Percona Server
● Support for all MySQL and Galera Cluster
● Remote-DBA Services for all MySQL
● MySQL Training
● Open Source Business Alliance (OSBA)
● Member of SOUG, DOAG, /ch/open
www.fromdual.com
3. www.fromdual.com
3 / 21
High-Availability :-(
● Who loves night-shifts?
● Who loves weekend-work?
● Who does regular upgrade (DB, kernel,
etc.)?
● Who does regular reboots (after kernel
upgrade)?
● Why are you not doing it in your office
hours?
6. www.fromdual.com
6 / 21
Advantages / Disadvantages
Based on InnoDB SE
Synchronous replication → No lost transaction
Active-active multi-master Cluster
→ Read and write to any cluster node (no r/w split any more!)
Read scalability and higher write throughput (Flash-Cache?)
Automatic node membership control
Rolling Restart (Upgrade of Hardware, O/S, DB release, etc.)
True parallel replication, on row level
→ No slave lag
A bit more complicated than normal MySQL, but similar
complexity as M/S Replication!
No original MySQL binaries → Codership MySQL binaries
Be aware of Hot Spots on rows: Higher probability of deadlocks
7. www.fromdual.com
7 / 21
Quorum and split-brain
● What is the problem?
● Split-brain → bad!
?
● Galera is a pessimistic Cluster → good!
● Quorum: FLOOR(n/2+1)
→ more than half! → 3-node Cluster (or 2+1)
14. www.fromdual.com
14 / 21
Initial Cluster start
● Start very 1st node with:
wsrep_cluster_address = "gcomm://"
or
mysqld_safe wsrepclusteraddress="
gcomm://"
● → this tells the node to be the first one!
● All other nodes normal:
service mysqld start
15. www.fromdual.com
15 / 21
Rolling Restart
● Scenario:
● Hardware-, O/S-, DB- and Galera-Upgrade
● MySQL configuration change
● During full operation!!! (99.999% HA, 5x9 HA)
● → Rolling Restart
● Start one node after the other in a cycle
● New features or settings are used after Rolling
Restart is completed
18. www.fromdual.com
18 / 21
Online Schema Upgrade (OSU)
● Schema Upgrade = DDL run against the DB
● Change DB structure
● Non transactional
● 2 Methods:
● Total Order Isolation (TOI) (default)
● Rolling Schema Upgrade (RSU)
● wsrep_osu_method = {TOI|RSU}
19. www.fromdual.com
19 / 21
Online Schema Upgrade
● Total Order Isolation (TOI) (default)
● Part of the database is locked for the duration of the DDL.
● + Simple, predictable and guaranteed data consistency.
● - Locking operation
● Good for fast DDL operations
● Rolling Schema Upgrade (RSU)
● DDL will be only processed locally at the node.
● Node is desynchronized for the duration of the DDL processing.
● After DDL completion, delayed write sets are applied (similar to IST).
● DDL should be manually executed at each node.
● + only blocking one node at a time
● - potentially unsafe and may fail if new and old schema are incompatible
● Good for slow DDL operations
20. www.fromdual.com
20 / 21
We want you!
● Database enthusiast for support /
remote-DBA / consulting
21. www.fromdual.com
21 / 21
Q & A
Questions ?
Discussion?
We have time for some face-to-face talks...
● FromDual provides neutral and independent:
● Consulting
● Remote-DBA
● Support for MySQL, Galera, Percona Server and MariaDB
● Training www.fromdual.com