Mark Cree left Hammerspace

Mark Cree, known since NuSpeed, Storspeed and InfiniteIO and finally AWS, with different trajectories has joined Hammerspace end of 2023 to lead strategic partnerships initiatives but he told me a few days ago that he left the company.

For our readers, NuSpeed, a pioneer in iSCSI storage router, has been acquired by Cisco in 2000 for $450M. Storspeed founded in 2007 has been swallowed by CacheIQ itself acquired by NetApp in 2010. We understand that Infinite.io, founded in 2012, has been closed without any transaction in 2021 and almost immediately Mark joined AWS.

He's now on the market for consulting mission advising storage companies.

Read More

The IT Press Tour #62 is back to California

The 62nd IT Press Tour will take place in a few days from June 2 to 6 in San Francisco and Silicon Valley.

Topics will be about IT infrastructure, cloud, networking, security, data management, big data, analytics and storage and of course AI as it is everywhere.

During this coming week, we'll meet 9 innovative companies and organizations:

• Cohesity, the new obvious global leader in Data Protection,
• DDN, the AI storage reference beyond its HPC storage leadership,
• Graid Technology, the pioneer in GPU-based RAID data protection,
• Hunch Tools, a emerging player developing AI tools,
• Lucidity, a fast growing FinOps player,
• Phison Electronics, the global leader in SSD and associated controller,
• PuppyGraph, a young analytics vendor promoting graph,
• Tabsdata, a new approach for data integration,
• and Ultra Accelerator Link Consortium, the authority for specifying interconnect between AI Accelerators.

I invite you to follow us on Twitter with #ITPT and @ITPressTour, my twitter handle and @CDP_FST and journalists' respective handle.

Read More

New VC round for Graid Technology

Graid Technology, the leader in GPU-based RAID data protection for NVMe storage, recently announced a new VC round, a series B with $30 million. This round shows a total of $48 million coming from HH-CTBC Partnership, a joint venture fund between Foxconn and CTBC, plus Yuanta Ventures and Delta Electronics Capital, Harbinger Venture Capital, associated with past investors from the Series A round in 2022. We'll learn more in a few days as Graid will join The IT Press Tour for the 3rd time.

Read More

DAOS, the exascale storage software

DAOS, acronym of Distributed Asynchronous Object Storage, is today an open source project supported by the DAOS Foundation. It powers some of the largest storage clusters on the planet. I invited Johann Lombardi, TSC chair of the foundation, during the recent IT Press Tour in London to learn more about the project and fix a few fuzzy comments and articles I see published on a few places.

Historically this large scale storage software and architecture has been initiated by Intel with the acquisition of Whamcloud in 2012. Intel asked Brent Gorda to lead that activity, he was the past CEO of Whamcloud. This move occurred after SUN acquisitions of CFS - Cluster File Systems - in September 2007 and ClusterStor by Xyratex in November 2010.

At that time, Intel had big storage ambitions with HPC, Big Data and Cloud flavors and some of you probably remembers the simple table, see below, I built in 2013 with several Intel's initiatives and investments in Cloudian, Amplidata, Maxta or Inktank in addition to Hadoop and Whamcloud moves.

Speaking about HPC and ExaScale, the below Intel picture also summarized pretty well the model at that time, it was circa mid 2013.

These elements are obviously linked to key people, execution and markets reality. And Whamcloud was later swallowed by DDN, it was in 2018.

In 2020, Intel surprised the market, at least part of it, and announced a joint-venture with SK Hynix for its NAND and SSD business to finally create Solidigm. Under the terms of this agreement, SK Hynix paid $7 billion and later $2 billion. Intel finally left this business, the market reality is that Flash and SSD business is controlled by Asian companies and US completely failed, Europe is absent by far as well.

The result of these various events culminated to the creation in 2023 of The DAOS Foundation, a new independent entity established to steward the ongoing development of DAOS as an open source project. This entity oversees governance, funding, and community building, ensuring DAOS remains vendor-neutral and sustainably developed. The founding members are Argonne National Laboratory, Google, HPE and Intel and Vdura, formerly Panasas, joined the organization.

Targeting high-end configurations with exascale requirements especially in terns of IO performance, DAOS implements some specific design choices. First, to clarify things, DAOS is not a file system, I read several times this, it is rather a storage software infrastructure that can be exposed and consumed by diverses interfaces, APIs and more globally access methods. But it implements some parallel mechanisms to boost performance as serialization is the enemy of performance. Among them, I wish to insist on IO operations that make design complex, adding latency that impacts the initial goal and mission of the platform. The teams has chosen to avoid read-modify-write operations, complex and increasing response time. It has been decided to pick a versioning model with a MVCC (multi-version concurrency control) to eliminate locking. Central metadata servers and global objet tables have been also removed to maximize scalability. As a horizontal model, DAOS works as a tight collaboration mode between server, in fact storage servers, and client layers with high numbers in mind, we speak here about thousands of nodes. On the server side, a DAOS engine runs connected with clients via specific user-space libraries such as libdfs and libdaos. 

Targeting high performance needs for research centers but also enterprise at large scale, DAOS offers a variety of interfaces with POSIX, NVMe-oF (block), MPI-IO, S3 and Hadoop. In addition, features like multi-tenancy, storage pools to segment data, key to share large configurations plus typed datasets.

Data is protected via replication and erasure coding delivering a high durability level and therefore resiliency. Erasure coding has the beauty to combine striping and redundancy without multi copies of data as only 1 copy is maintained coupled with redundant chunks, fragments or segments depending of the terminology used.

DAOS has demonstrated its performance level being adopted by some of the largest and fastest supercomputers on the planet. This is illustrated by the IO500 benchmark list and the Aurora system, ranked #1, deployed at Argonne National Laboratory.

The foundation plans to announce and release the 2.8 version during SC25 organized in November in St Louis, MI. In the meantime, DAOS will be present at ISC in Hambourg mid June as several vendors will exhibit there. So more news soon.

Read More

Hammerspace raised $100M

Hammerspace, a fast growing company playing in the HPC and more recently in AI storage infrastructure, just announced an impressive VC round with $100 million for its Series B. The total raised approaches $157M.

As AI changed lots of things, the company has changed its key message claiming to the high-performance data platform for AI. It's a good indicator for Hammerspace, now sales execution is a must.

I remind our readers that Hammerspace has roots in Primary Data being founded with the acquisition of Tonian Systems in 2013, they also acquired Rozo Systems, an confidential acquisition I unveiled to the world with this article. We'll monitor carefully the trajectory.

Read More

Quesma unveils in preview its charts

Quesma, a developer of a SQL database gateway, took advantage of its participation of The IT Press Tour in London recently to unveil their Charts service. Surfing on AI and various services it can deliver, Jacek Migdal, CEO, gave us in exclusivity a demo. It is an AI oriented service that generates charts, graphics, histograms on various statistics or historical society, sport... numbers from CSV, Excel or SQL data. Give it a try, you can subscribe to the service at https://charts.quesma.com/.

Read More

FerretDB promotes an alternative to MongoDB

MongoDB is hot and appears as a monopoly for Document DB, one of the NoSQL database types available on the market. Founded in 2007, MongoDB, started as 10Gen, developed and promoted an open source approach. With the arrival of Dev Ittycheria as CEO in 2014, the shift to a real business machine was initiated and the company adopted a close model and terminated its open source philosophy. The company did its IPO in 2017 at Nasdaq under the ticket MDB and recent market capitalization reaches $13B. The annual revenue is around $2B for its last financial year. But it turns out that users, who have been attracted by a clear obvious value both from a product perspective and license model, were impacted and some alternatives were triggered.

One of the alternatives is FerretDB, a project started by a group of talented people such as Peter Farkas, CEO, Alexey Palazhchenko, CTO, 2 ex-Percona guys, and Peter Zaitsev, advisor, former CEO of Percona. The firm, based in Delaware, US, has raised $2.8M in 2022 and currently has 8 people. 

We had the chance to meet Peter during the recent IT Press Tour organized in London early April. And I'm particularly glad that Peter joined us as I tried to invite him for several months.

FerretDB has its roots in the original open source model MongoDB used and its shift to close source in 2018 plus issues it created at customers sites with vendor lock-in. Probably not on the technical aspect but rather on the financial dimension. In detail MongoDB uses the SSPL - Server Side Public License - and developed a specific query language, MQL for MongoDB Query Language, that has even more locked users.

So what is FerretDB?

In simple words, FerretDB is a compatible MongoDB database built on PostgreSQL and distributed in open source.

Starting the FerretDB project at the end of 2021, MongoDB tried to stop the initiative with legal pressure but it appeared that it went nowhere so far. The FerretDB team has designed, and written in Go, a layer on top of PostgreSQL database and runs on-premise or in the cloud under the Apache 2.0 license.

The goal is to run an application built with MongoDB MQL to operate the same way without any changes on top of FerretDB.  The architecture they adopted, described below, illustrates the approach perfectly.

Today, more than 4,000 instances are running with FerretDB but there should be much more as these are only the ones with Telemetry enabled. Even with 8 people, the project has more than 200 contributors and received 10k stars on GitHub.

In 2023, FerretDB initiated a collaboration with Microsoft to address the MongoDB challenge.

This story justifies the need for an Open Standard for Document Databases in order to democratize the model.

Clearly, the industry needs projects like FerretDB. Good luck guys.

Read More

Quesma promotes database gateways

Quesma, a young Polish software company, joined the recent IT Press Tour in London, organized in parallel with KubeCon. We first discussed with Quesma during the Boston edition last October when we met Hydrolix as Quesma provides a key core component for their solution. And it gave me ideas to learn more about their solution.

We had the opportunity to meet Jacek Migdal, CEO and founder, for a company and product update. Quesma, founded around the end of 2023, raised $2.5 million and has 8 employees with a deep expertise in structured data, SQL language and database back-ends. And even if many tries have been made to replace SQL, the query language is here to stay as it is so widely used and adopted. But what about the back-end databases that could be replaced with more modern approaches and designs especially in log, traces, events and metrics world.

The team develops a middleware, operating as a database gateway to proxy queries, in other words, applications won't talk directly with back-end databases but rather are connected to Quesma query service layer. The solution has been recognized and adapted by 2 hot players, Hydrolix and ClickHouse. In terms of use case, migration is of course addressed but also database integration without changing anything on the application layer.

The team prepares SQL Pipe Syntax based on the Google paper available here. The idea is to optimize SQL commands and leverage some semantic order alignments.

Quesma product is available on GitHub and the pricing is aligned to the stack used by users. In other words, if the stack is open source, Quesma is free, if not a fee has to be paid at $30/TB/month.

Read More

Hydrolix raised $80M Series C

Hydrolix, a fast growing player in the streaming data lake landscape, just announced its new VC rounds, a series C of $80 million for a total of $145 million. The company continues to grow at a rapid scale multiplying the revenue by 8 between the 2 rounds i.e in less than 1 year with also a significant success of the Akamai deal named as TrafficPeak. In 2024, the solution has been used for the Super Bowl, the Olympics Game and Black Friday monitoring at a massive scale. We met the company in Boston during the 58th edition of The IT Press Tour. This is also the session where we discovered Quesma, a Polish company paramount in the Hydrolix success for back-end database connection, that we just met in London during the 61st tour with its CEO and co-founder Jacek Migdał. 

Read More

The IT Press Tour #61 will take place at the famous Battersea Power Station

The 61st IT Press Tour will take place next week, April 1st & 2nd, in London, England, exactly at the Battersea Power Station. This area is famous with the Pink Floyd Animals album cover issued in 1977.

Topics will be about IT infrastructure, cloud, networking, security, data management, big data, analytics and storage and of course AI as it is everywhere.

As usual with the European tour format, we'll meet 6 innovative companies over these 2 days:

• AppsCode, a young Kubernetes IT operations player,
• Auwau, a recent cloud backup operations management vendor,
• DAOS, a reference in hyper performance storage infrastructure,
• FerretDB, a emerging player developing a compatible MongoDB engine,
• Quesma, a specialist in portable SQL across applications,
• and Storadera, the alternative to AWS S3 and Wasabi.

I invite you to follow us on Twitter with #ITPT and @ITPressTour, my twitter handle and @CDP_FST and journalists' respective handle.

Read More