Jun
11
2020
--

OpenStack adds the StarlingX edge computing stack to its top-level projects

The OpenStack Foundation today announced that StarlingX, a container-based system for running edge deployments, is now a top-level project. With this, it joins the main OpenStack private and public cloud infrastructure project, the Airship lifecycle management system, Kata Containers and the Zuul CI/CD platform.

What makes StarlingX a bit different from some of these other projects is that it is a full stack for edge deployments — and in that respect, it’s maybe more akin to OpenStack than the other projects in the foundation’s stable. It uses open-source components from the Ceph storage platform, the KVM virtualization solution, Kubernetes and, of course, OpenStack and Linux. The promise here is that StarlingX can provide users with an easy way to deploy container and VM workloads to the edge, all while being scalable, lightweight and providing low-latency access to the services hosted on the platform.

Early StarlingX adopters include China UnionPay, China Unicom and T-Systems. The original codebase was contributed to the foundation by Intel and Wind River System in 2018. Since then, the project has seen 7,108 commits from 211 authors.

“The StarlingX community has made great progress in the last two years, not only in building great open source software but also in building a productive and diverse community of contributors,” said Ildiko Vancsa, ecosystem technical lead at the OpenStack Foundation. “The core platform for low-latency and high-performance applications has been enhanced with a container-based, distributed cloud architecture, secure booting, TPM device enablement, certificate management and container isolation. StarlingX 4.0, slated for release later this year, will feature enhancements such as support for Kata Containers as a container runtime, integration of the Ussuri version of OpenStack, and containerization of the remaining platform services.”

It’s worth remembering that the OpenStack Foundation has gone through a few changes in recent years. The most important of these is that it is now taking on other open-source infrastructure projects that are not part of the core OpenStack project but are strategically aligned with the organization’s mission. The first of these to graduate out of the pilot project phase and become top-level projects were Kata Containers and Zuul in April 2019, with Airship joining them in October.

Currently, the only pilot project for the OpenStack Foundation is its OpenInfra Labs project, a community of commercial vendors and academic institutions, including the likes of Boston University, Harvard, MIT, Intel and Red Hat, that are looking at how to better test open-source code in production-like environments.

 

May
04
2020
--

IBM and Red Hat expand their telco, edge and AI enterprise offerings

At its Think Digital conference, IBM and Red Hat today announced a number of new services that all center around 5G edge and AI. The fact that the company is focusing on these two areas doesn’t come as a surprise, given that both edge and AI are two of the fastest-growing businesses in enterprise computing. Virtually every telecom company is now looking at how to best capitalize on the upcoming 5G rollouts, and most forward-looking enterprises are trying to figure out how to best plan around this for their own needs.

As IBM’s recently minted president Jim Whitehurst told me ahead of today’s announcement, he believes that IBM (in combination with Red Hat) is able to offer enterprises a very differentiated service because, unlike the large hyper clouds, IBM isn’t interested in locking these companies into a homogeneous cloud.

“Where IBM is competitively differentiated, is around how we think about helping clients on a journey to what we call hybrid cloud,” said Whitehurst, who hasn’t done a lot of media interviews since he took the new role, which still includes managing Red Hat. “Honestly, everybody has hybrid clouds. I wish we had a more differentiated term. One of the things that’s different is how we’re talking about how you think about an application portfolio that, by necessity, you’re going to have in multiple ways. If you’re a large enterprise, you probably have a mainframe running a set of transactional workloads that probably are going to stay there for a long time because there’s not a great alternative. And there’s going to be a set of applications you’re going to want to run in a distributed environment that need to access that data — all the way out to you running a factory floor and you want to make sure that the paint sprayer doesn’t have any defects while it’s painting a door.”

BARCELONA, CATALONIA, SPAIN – 2019/02/25: The IBM logo is seen during MWC 2019. (Photo by Paco Freire/SOPA Images/LightRocket via Getty Images)

He argues that IBM, at its core, is all about helping enterprises think about how to best run their workloads software, hardware and services perspective. “Public clouds are phenomenal, but they are exposing a set of services in a homogeneous way to enterprises,” he noted, while he argues that IBM is trying to weave all of these different pieces together.

Later in our discussion, he argued that the large public clouds essentially force enterprises to fit their workloads to those clouds’ service. “The public clouds do extraordinary things and they’re great partners of ours, but their primary business is creating these homogeneous services, at massive volumes, and saying ‘if your workloads fit into this, we can run it better, faster, cheaper etc.’ And they have obviously expanded out. They’ve added services. They are not saying we can put a box on-premise, but you’re still fitting into their model.”

On the news side, IBM is launching new services to automate business planning, budgeting and forecasting, for example, as well as new AI-driven tools for building and running automation apps that can handle routine tasks either autonomously or with the help of a human counterpart. The company is also launching new tools for call-center automation.

The most important AI announcement is surely Watson AIOps, though, which is meant to help enterprises detect, diagnose and respond to IT anomalies in order to reduce the effects of incidents and outages for a company.

On the telco side, IBM is launching new tools like the Edge Application Manager, for example, to make it easier to enable AI, analytics and IoT workloads on the edge, powered by IBM’s open-source Open Horizon edge computing project. The company is also launching a new Telco Network Cloud manager built on top of Red Hat OpenShift and the ability to also leverage the Red Hat OpenStack Platform (which remains to be an important platform for telcos and represents a growing business for IBM/Red Hat). In addition, IBM is launching a new dedicated IBM Services team for edge computing and telco cloud to help these customers build out their 5G and edge-enabled solutions.

Telcos are also betting big on a lot of different open-source technologies that often form the core of their 5G and edge deployments. Red Hat was already a major player in this space, but the acquisition has only accelerated this, Whitehurst argued. “Since the acquisition […] telcos have a lot more confidence in IBM’s capabilities to serve them long term and be able to serve them in mission-critical context. But importantly, IBM also has the capability to actually make it real now.”

A lot of the new telco edge and hybrid cloud deployments, he also noted, are built on Red Hat technologies but built by IBM, and neither IBM nor Red Hat could have really brought these to fruition in the same way. Red Hat never had the size, breadth and skills to pull off some of these projects, Whitehurst argued.

Whitehurst also argued that part of the Red Hat DNA that he’s bringing to the table now is helping IBM to think more in terms of ecosystems. “The DNA that I think matters a lot that Red Hat brings to the table with IBM — and I think IBM is adopting and we’re running with it — is the importance of ecosystems,” he said. “All of Red Hat’s software is open source. And so really, what you’re bringing to the table is ecosystems.”

It’s maybe no surprise then that the telco initiatives are backed by partners like Cisco, Dell Technologies, Juniper, Intel, Nvidia, Samsung, Packet, Equinix, Hazelcast, Sysdig, Turbonomics, Portworx, Humio, Indra Minsait, EuroTech, Arrow, ADLINK, Acromove, Geniatech, SmartCone, CloudHedge, Altiostar, Metaswitch, F5 Networks and ADVA.

In many ways, Red Hat pioneered the open-source business model and Whitehurst argued that having Red Hat as part of the IBM family means it’s now easier for the company to make the decision to invest even more in open source. “As we accelerate into this hybrid cloud world, we’re going to do our best to leverage open-source technologies to make them real,” he added.

May
04
2020
--

Nvidia acquires Cumulus Networks

Nvidia today announced its plans to acquire Cumulus Networks, an open-source-centric company that specializes in helping enterprises optimize their data center networking stack. Cumulus offers both its own Linux distribution for network switches, as well as tools for managing network operations. With Cumulus Express, the company also offers a hardware solution in the form of its own data center switch.

The two companies did not announce the price of the acquisition, but chances are we are talking about a considerable amount, given that Cumulus had raised $134 million since it was founded in 2010.

Mountain View-based Cumulus already had a previous partnership with Mellanox, which Nvidia acquired for $6.9 billion. That acquisition closed only a few days ago. As Mellanox’s Amit Katz notes in today’s announcement, the two companies first met in 2013, and they formed a first official partnership in 2016. Cumulus, it’s worth noting, was also an early player in the OpenStack ecosystem.

Having both Cumulus and Mellanox in its stable will give Nvidia virtually all the tools it needs to help enterprises and cloud providers build out their high-performance computing and AI workloads in their data centers. While you may mostly think about Nvidia because of its graphics cards, the company has a sizable data center group, which delivered close to $1 billion in revenue in the last quarter, up 43% from a year ago. In comparison, Nvidia’s revenue from gaming was just under $1.5 billion.

“With Cumulus, NVIDIA can innovate and optimize across the entire networking stack from chips and systems to software including analytics like Cumulus NetQ, delivering great performance and value to customers,” writes Katz. “This open networking platform is extensible and allows enterprise and cloud-scale data centers full control over their operations.”

Apr
22
2020
--

Granulate announces $12M Series A to optimize infrastructure performance

As companies increasingly look to find ways to cut costs, Granulate, an early-stage Israeli startup, has come up with a clever way to optimize infrastructure usage. Today it was rewarded with a tidy $12 million Series A investment.

Insight Partners led the round with participation from TLV Partners and Hetz Ventures. Lonne Jaffe, managing director at Insight Partners, will be joining the Granulate board under the terms of the agreement. Today’s investment brings the total raised to $15.6 million, according to the company.

The startup claims it can cut infrastructure costs, whether on-prem or in the cloud, from between 20% and 80%. This is not insignificant if they can pull this off, especially in the economic maelstrom in which we find ourselves.

Asaf Ezra, co-founder and CEO at Granulate, says the company achieved the efficiency through a lot of studying about how Linux virtual machines work. Over six months of experimentation, they simply moved the bottleneck around until they learned how to take advantage of the way the Linux kernel operates to gain massive efficiencies.

It turns out that Linux has been optimized for resource fairness, but Granulate’s founders wanted to flip this idea on its head and look for repetitiveness, concentrating on one function instead of fair allocation across many functions, some of which might not really need access at any given moment.

“When it comes to production systems, you have a lot of repetitiveness in the machine, and you basically want it to do one thing really well,” he said.

He points out that it doesn’t even have to be a VM. It could also be a container or a pod in Kubernetes. The important thing to remember is that you no longer care about the interactivity and fairness inherent in Linux; instead, you want that the machine to be optimized for certain things.

“You let us know what your utility function for that production system is, then our agents. basically optimize all the decision making for that utility function. That means that you don’t even have to do any code changes to gain the benefit,” Ezra explained.

What’s more, the solution uses machine learning to help understand how the different utility functions work to provide greater optimization to improve performance even more over time.

Insight’s Jaffe certainly recognized the potential of such a solution, especially right now.

“The need to have high-performance digital experiences and lower infrastructure costs has never been more important, and Granulate has a highly differentiated offering powered by machine learning that’s not dependent on configuration management or cloud resource purchasing solutions,” Jaffe said in a statement.

Ezra understands that a product like his could be particularly helpful at the moment. “We’re in a unique position. Our offering right now helps organizations survive the downturn by saving costs without firing people,” he said.

The company was founded in 2018 and currently has 20 employees. They plan to double that by the end of 2020.

Apr
06
2020
--

Paul Cormier takes over as Red Hat CEO, as Jim Whitehurst moves to IBM

When Ginni Rometty indicated that she was stepping down as IBM CEO at the end of January, the company announced that Arvind Krishna would be taking over, while Red Hat CEO Jim Whitehurst would become president. To fill his role, Red Hat announced today that long-time executive Paul Cormier has been named president and CEO.

Cormier would seem to be a logical choice to run Red Hat, having been with the company since 2001. He joined as its VP of engineering and has seen the company grow from a small startup to a multi-billion dollar company.

Cormier spoke about the historical arc he has witnessed in his years at Red Hat. “Looking back to when I joined, we were in a different position and facing different issues, but the spirit was the same. We were on a mission to convince the world that open source was real, safe and enterprise-grade,” Cormier said in an email to employees about his promotion.

Former CEO Whitehurst certainly sees this as a sensible transition. “After working with him closely for more than a decade, I can confidently say that Paul was the natural choice to lead Red Hat. Having been the driving force behind Red Hat’s product strategy for nearly two decades, he’s been intimately involved in setting the company’s direction and uniquely understands how to help customers and partners make the most out of their cloud strategy,” he said in a statement.

In a Q&A with Cormier on the company website, he talked about the kind of changes he expects to see under his leadership in the next five years of the company. “There’s a term that we use today, ‘applications run the business.’ In five years, I see it becoming the case for the majority of enterprises. And with that, the infrastructure underpinning these applications will be even more critical. Management and security are paramount — and this isn’t just one environment. It’s bare metal and hypervisors to public and private clouds. It’s Linux, VMs, containers, microservices and more,” he said.

When IBM bought Red Hat in 2018 for $34 billion, there was widespread speculation that Whitehurst would eventually take over in an executive position there. Now that that has happened, Cormier will step into run Red Hat.

While Red Hat is under the IBM umbrella, it continues to operate as a separate company with its own executive structure, but that vision that Cormier outlined is in line with how it will fit within the IBM family as it tries to make its mark on the shifting cloud and enterprise open source markets.

Mar
11
2020
--

AWS launches Bottlerocket, a Linux-based OS for container hosting

AWS has launched its own open-source operating system for running containers on both virtual machines and bare metal hosts. Bottlerocket, as the new OS is called, is basically a stripped-down Linux distribution that’s akin to projects like CoreOS’s now-defunct Container Linux and Google’s container-optimized OS. The OS is currently in its developer preview phase, but you can test it as an Amazon Machine Image for EC2 (and by extension, under Amazon EKS, too).

As AWS chief evangelist Jeff Barr notes in his announcement, Bottlerocket supports Docker images and images that conform to the Open Container Initiative image format, which means it’ll basically run all Linux-based containers you can throw at it.

One feature that makes Bottlerocket stand out is that it does away with a package-based update system. Instead, it uses an image-based model that, as Barr notes, “allows for a rapid & complete rollback if necessary.” The idea here is that this makes updates easier. At the core of this update process is “The Update Framework,” an open-source project hosted by the Cloud Native Computing Foundation.

AWS says it will provide three years of support (after General Availability) for its own builds of Bottlerocket. As of now, the project is very much focused on AWS, of course, but the code is available on GitHub and chances are we will see others expand on AWS’ work.

The company is launching the project in cooperation with a number of partners, including Alcide, Armory, CrowdStrike, Datadog, New Relic, Sysdig, Tigera, Trend Micro and Waveworks.

“Container-optimized operating systems will give dev teams the additional speed and efficiency to run higher throughput workloads with better security and uptime,” said Michael Gerstenhaber, director of Product Management at Datadog.” We are excited to work with AWS on Bottlerocket, so that as customers take advantage of the increased scale they can continue to monitor these ephemeral environments with confidence.”

 

Feb
24
2020
--

Databricks makes bringing data into its ‘lakehouse’ easier

Databricks today announced the launch of its new Data Ingestion Network of partners and the launch of its Databricks Ingest service. The idea here is to make it easier for businesses to combine the best of data warehouses and data lakes into a single platform — a concept Databricks likes to call “lakehouse.”

At the core of the company’s lakehouse is Delta Lake, Databricks’ Linux Foundation-managed open-source project that brings a new storage layer to data lakes that helps users manage the lifecycle of their data and ensures data quality through schema enforcement, log records and more. Databricks users can now work with the first five partners in the Ingestion Network — Fivetran, Qlik, Infoworks, StreamSets, Syncsort — to automatically load their data into Delta Lake. To ingest data from these partners, Databricks customers don’t have to set up any triggers or schedules — instead, data automatically flows into Delta Lake.

“Until now, companies have been forced to split up their data into traditional structured data and big data, and use them separately for BI and ML use cases. This results in siloed data in data lakes and data warehouses, slow processing and partial results that are too delayed or too incomplete to be effectively utilized,” says Ali Ghodsi, co-founder and CEO of Databricks. “This is one of the many drivers behind the shift to a Lakehouse paradigm, which aspires to combine the reliability of data warehouses with the scale of data lakes to support every kind of use case. In order for this architecture to work well, it needs to be easy for every type of data to be pulled in. Databricks Ingest is an important step in making that possible.”

Databricks VP of Product Marketing Bharath Gowda also tells me that this will make it easier for businesses to perform analytics on their most recent data and hence be more responsive when new information comes in. He also noted that users will be able to better leverage their structured and unstructured data for building better machine learning models, as well as to perform more traditional analytics on all of their data instead of just a small slice that’s available in their data warehouse.

Jan
21
2020
--

Canonical’s Anbox Cloud puts Android in the cloud

Canonical, the company behind the popular Ubuntu Linux distribution, today announced the launch of Anbox Cloud, a new platform that allows enterprises to run Android in the cloud.

On Anbox Cloud, Android becomes the guest operating system that runs containerized applications. This opens up a range of use cases, ranging from bespoke enterprise apps to cloud gaming solutions.

The result is similar to what Google does with Android apps on Chrome OS, though the implementation is quite different and is based on the LXD container manager, as well as a number of Canonical projects like Juju and MAAS for provisioning the containers and automating the deployment. “LXD containers are lightweight, resulting in at least twice the container density compared to Android emulation in virtual machines – depending on streaming quality and/or workload complexity,” the company points out in its announcements.

Anbox itself, it’s worth noting, is an open-source project that came out of Canonical and the wider Ubuntu ecosystem. Launched by Canonical engineer Simon Fels in 2017, Anbox runs the full Android system in a container, which in turn allows you to run Android application on any Linux-based platform.

What’s the point of all of this? Canonical argues that it allows enterprises to offload mobile workloads to the cloud and then stream those applications to their employees’ mobile devices. But Canonical is also betting on 5G to enable more use cases, less because of the available bandwidth but more because of the low latencies it enables.

“Driven by emerging 5G networks and edge computing, millions of users will benefit from access to ultra-rich, on-demand Android applications on a platform of their choice,” said Stephan Fabel, director of Product at Canonical, in today’s announcement. “Enterprises are now empowered to deliver high performance, high density computing to any device remotely, with reduced power consumption and in an economical manner.”

Outside of the enterprise, one of the use cases that Canonical seems to be focusing on is gaming and game streaming. A server in the cloud is generally more powerful than a smartphone, after all, though that gap is closing.

Canonical also cites app testing as another use case, given that the platform would allow developers to test apps on thousands of Android devices in parallel. Most developers, though, prefer to test their apps in real — not emulated — devices, given the fragmentation of the Android ecosystem.

Anbox Cloud can run in the public cloud, though Canonical is specifically partnering with edge computing specialist Packet to host it on the edge or on-premise. Silicon partners for the project are Ampere and Intel .

Dec
10
2019
--

Microsoft announces public preview of Microsoft Teams for Linux

Today, Microsoft announced a public preview of Microsoft Teams for Linux, the first Office 365 tool that’s available for the open-source operating system.

The hope is that by making it available for preview, the company can get feedback from the community and improve it before it becomes generally available. “Starting today, Microsoft Teams is available for Linux users in public preview, enabling high quality collaboration experiences for the open source community at work and in educational institutions,” the company wrote in the blog post announcing the release.

The goal here ultimately is to help get Teams into the hands of more customers by expanding the platforms it runs on. “Most of our customers have devices running on a variety of different platforms such as Windows 10, Linux and others. We are committed to supporting mixed environments across our cloud and productivity offerings, and with this announcement, we are pleased to extend the Teams experience to Linux users,” the company wrote in the blog post.

This announcement is significant for a couple of reasons. For starters, Microsoft has had a complicated history with Linux and open source, although in recent years, under Satya Nadella, it has embraced open source. This shows that Microsoft is willing to put its tools wherever customers need them, regardless of the platform or operating system.

Secondly, as it marks the first Office 365 app on Linux, if there is positive feedback, it could open the door for more apps on the platform down the road.

The announcement also comes against the backdrop of the company’s ongoing battles with Slack for enterprise collaboration platform users. In July, Microsoft announced 13 million daily active users on Teams. Meanwhile, Slack has 12 million DAUs. It’s worth noting that Slack has been available on Linux for almost two years.

Nov
12
2019
--

Watch Out for Disk I/O Performance Issues when Running EXT4

Performance Issues When Running EXT4

Recently, at Percona Live Europe 2019, Dimitri Kravchuk from Oracle mentioned that he observed some unclear drop in performance for MySQL on an ext4 filesystem with the latest Linux kernels. I decided to check this case out on my side and found out that indeed, starting from linux kernel 4.9, there are some cases with notable (up to 2x) performance drops for ext4 filesystem in direct i/o mode.

So what’s wrong with ext4? It started in 2016 from the patch that was pushed to kernel 4.9: “ext4: Allow parallel DIO reads”. The purpose of that patch was to help to improve read scalability in direct i/o mode. However, along with improvements in pure read workloads, it also introduced regression in intense mixed random read/write scenarios. And it’s quite weird, but this issue had not been noticed for 3 years. Only this summer was performance regression reported and discussed in LKML. As a result of this discussion, there is an attempt to fix it, but from my current understanding that fix will be pushed only to upcoming 5.4/5.5 kernels. Below I will describe what this regression looks like, how it affects MySQL workloads, and what workarounds we can apply to mitigate this issue.

ext4 Performance Regression

Let’s start by defining the scope of this ext4 performance regression. It will only have an impact if the setup/workload meets following conditions:
– fast ssd/nvme
– linux kernel>=4.9
– files resides on ext4 file system
– files opened with O_DIRECT flag
– at least some I/O should be synchronous

In the original report to LKML, the issue was observed/reproduced with a mixed random read/write scenario with sync I/O and O_DIRECT. But how do these factors relate to MySQL? The only files opened by InnoDB in O_DIRECT mode are tablespaces (*.ibd files), and I/O pattern for tablespaces consists of following operations:

– reads ibd data in synchronous mode
– writes ibd data in asynchronous mode
– posix_allocate to extend tablespace file followed by a synchronous write
– fsync

There are also extra I/O from WAL log files:

– writes data to log files in synchronous mode
– fsync

So in the case of InnoDB tablespaces that are opened with O_DIRECT, we have a mix of sync reads and async writes and it turned out that such a combination along with sync writes to innodb log file is enough to cause notable performance regression as well. I have sketched the workload for fio tool (see below) that simulates the I/O access pattern for InnoDB and have run it for SSD and NVMe drives for linux kernels 4.4.0, 5.3.0, and 5.3.0 with ext4 scalability fix.

[global]
filename=tablespace1.ibd:tablespace2.ibd:tablespace3.ibd:tablespace4.ibd:tablespace5.ibd
direct=1
bs=16k
iodepth=1

#read data from *.ibd tablespaces
[ibd_sync_read]
rw=randread
ioengine=psync

#write data to *.ibd tavlespaces
[ibd_async_write]
rw=randwrite
ioengine=libaio

#write data to ib* log file
[ib_log_sync_write]
rw=write
bs=8k
direct=0
ioengine=psync
fsync=1
filename=log.ib
numjobs=1

fio results on the chart:

Observations:

– for SATA/SSD drive there is almost no difference in throughtput, and only at 16 threads do we see a drop in reads for ext4/kernel-5.3.0. For ext4/kernel-5.3.0 mounted with dioread_nolock (that enables scalability fixes), we see that IOPS back and even look better.
– for NVMe drive the situation looks quite different – until 8 i/o threads IOPS for both reads and writes are more/less similar but after increasing pressure on i/o we see a notable spike for writes and similar drop for reads. And again mounting ext4 with dioread_nolock helps to get the same throughput as and for kernels < 4.9.

The similar performance data for the original issue reported to LKML (with more details and analysis) can be found in the patch itself.

How it Affects MySQL

O_DIRECT

Now let’s check the impact of this issue on an IO-bound sysbench/OLTP_RW workload in O_DIRECT mode. I ran a test for the following setup:

– filesystem: xfs, ext4/default, ext4/dioread_nolock
– drives: SATA/SSD and NVMe
– kernels: 4.4.0, 5.3.0, 5.3.0+ilock_fix

Observations

– in the case of SATA/SSD, the ext4 scalability issue has an impact on tps rate after 256 threads and drop is 10-15%
– in the case of NVMe and regular ext4 with kernel 5.3.0 causes performance drop in ~30-80%. If we apply a fix by mounting ext4 with dioread_nolock or use xfs,  throughput looks good.

O_DSYNC

As ext4 regression affects O_DIRECT, let’s replace O_DIRECT with O_DSYNC and look at results of the same sysbench/OLTP_RW workload on kernel 5.3.0:

Note: In order to make results between O_DIRECT and O_DSYNC comparable, I have limited available memory for MySQL instance by cgroup.

Observations:

In the case of O_DSYNC and regular ext4, the performance is just 10% less than for O_DIRECT/ext4/dioread_nolock and O_DIRECT/xfs and ~35% better than for O_DIRECT/ext4. That means that O_DSYNC can be used as a workaround for cases when you have fast storage and ext4 as filesystem but can’t switch to xfs or upgrade kernel.

Conclusions/workarounds

If your workload/setup is affected, there are the following options that you may consider as a workaround:

– downgrade linux kernel to 4.8
– install kernel 5.3.0 with fix and mount ext4 with dioread_nolock option
– if O_DIRECT is important, switch to xfs filesystem
– if changing filesystem is not an option,  replace O_DIRECT with O_DSYNC+cgroup

Powered by WordPress | Theme: Aeros 2.0 by TheBuckmaker.com