Posts by josh.krischer

Fresh cotton underpants fix series of mysterious mainframe crashes

I had a similar case in the 70-ies:

Almost every day an IBM 360/65 was crashing at 7:30 +,- few minutes. Nothing in log, any other indication. My usual starting time was 08:00 but I decided to start earlier and to watch the machine. At 7:30 the cleaning lady moved with her trolley near the gate which contained the core memory of the mainframe and the system crashed.Next day I asked her to take another route and nothing happen. Most likely the reason was static electricity developed by the wheels of the trolley.

EMC publicly denies Fairview borkage involvement

Did you ever hear EMC admitting a failure? It is always the customer mistake.

EMC+Dell: Firm made a $2bn bed-hopping proposal

this agreement reminds me a contract between pig and chicken to produce bacon and eggs. As usual EMC is the chicken.

Big Blue drops a billion dollars to refresh its pig herd's lipstick

IBM - the SDS leader?

Currently most of the storage Control Units are based on the x86 processors, same technology as the servers; multi core Intel chips. In fact the multi core is used much more effectively is CU than in servers. Multi-core technology and server virtualization bring some other developments to watch such as the Virtual SAN Appliances (VSAs) or Software Defined Storage SDS) and embedded application on storage control units. The first emulates the server as storage CU, the second using storage CU for applications. SDS is one of the top considerations of IT executives. After selling the x Series to Lenovo, IBM margins on the server hardware are minimal .IBM simply disconnected between the hardware and the control program of almost all storage devices based on x86 platform. It means that the users can use unused or 3rd party servers (or partitions under VMware) to become storage control units, which in many cases can save money. However, it is for the users to deploy servers according to IBM specification and to take responsibility on their availability.

This step, practically makes IBM the leader in SDS offering such solution for tapes, virtualization, XIV storage and clustered NAS. Not a single other storage vendor is coming close to this palette.

HDS, HP leapfrog NetApp in the land of the MAGIC QUADRANT

@Rob - I agree, what I meant was among the hi-end storage subsystems; DS8xxx, VMX and VSP.

Thanks pd_p

Bundling VMAX, VNX , Isilon, ScaleIO, DSSD, etc. in one MQ report may be good for investors but not for the end-users. Different products of the same vendor show different visions, in particular in EMC which use very little synergy among different product lines.

I have many reservations to the positioning and explanations but will mention only two about IBM positioning. For Example

MQ quote:"A steady cadence of VMAX, VNX and Isilon product enhancements, new model announcements, software improvements and strategy white papers keep market attention focused on these products."

JK: The IBM DS8000 pace of development is faster than the competition. In the last 10 years IBM launched seven models of the DS8xxx, more than a double in comparison to the competition. At least twice a year IBM introduces a new model, new features or both.

MQ quote : "The lack of timely responses to competitors' new product announcements, coupled with IBM's termination of its OEM relationship with NetApp and its divestiture of its server business, has created user concerns about IBM's commitment to the disk storage business".

JK: IBM was the first to introduce sub LUN automated tiering, The Easy Tier includes the Fusion I/O cards in the POWER server, Heat Map Transfer between the DS8000 in remote mirroring or the High Performance Flash Enclosure on the PCI bus of the DS8870- all these are unique features. On the StorWize series, the virtualization, real-time compression with Hardware Compression Accelerators,etc. The XIV GUI which is common GUI.

Chris, I am not sure that this MQ is for high-end VMAX, VSP, DS8870 at all. Gartner, after the launch of the HDS VSP in 2004 replaced the MQ of the hi-end with Market Scope.

Anyhow, as a past co-author of the MQs, some of the positions on this MQ are questionable.

IBM handing pink slips to Israel-based Diligent devs - reports

To Anonymous Coward

The name fits you. Maybe Ignorant Anonymous Coward fits you better.

Dell spits up storage revenues after gobbling promising upstarts

Most of Dell customers do not require high-end storage. When Sun started resell Hitachi the situation was different, most of the storage on the top 10 000 series of Sun was EMC Symmetrix. Both Compellent and EqualLogics are innovative products (Compellent for example was the first to introduce automated data placement, etc.) which fit Dell much better than large high-end subsystems. Comparing Server/storage ratio you have to consider the current weakness of hp which Dell enjoy.

They could do better with integration of all the companies which the acquired but at the end the problem of Dell in storage is not products but execution.

Is that a flash stick in your pocket, VMware, or are you pleased to see us?

@K. The whole idea of SDS as VSA is saving storage CU by using JBODs or internal media. The multi-core processors with storage CU software running in partition may deliver limited functionality with good enough performance for SMEs with lower investments in material costs. Software can be multiplicated indefinitely so the cost s/b reasonable.

Summary

Emulating shared block-access storage area network (SAN) system on internal DAS storage (running in virtualization partition)

The server SW acts as SAN array controller software

A pair of servers with high amount of direct-accessed storage (DAS), accessed by other servers across a network

LeftHand Networks pioneered, evolved as HPs StoreVirtual (VSA), NetApp’s ONTAP Edge, Nexenta, StorMagic (SvSAN), Nutanix , ScaleIO (acq. by EMC)

Mellanox Storage Accelerator VSA product accessed over Ethernet or InfiniBand supports DAS & SAN promise better performance

Saves HBAs, switches, Physical CU price

No one ever got fired for buying enterprise storage, right?

High-End Enterprise Array:

Functionality, availability

More processing power to support functionality

Multiple LUN access, load balancing

Greater scalability

Connectivity, platforms

Capacity

Simplified Management

-Capacity

-Disaster recovery

-T-uning

Data-center-level serviceand support

Non-disruptive upgrades and repairs, microcode change

Full redundancy

Small impact of a component failure

Enterprise-wide

zOS, RISC Unix, Windows, Linux

Data integrity and data consistency techniques across multiple subsystem

Midrange Enterprise Array:

Lower price

Risk dispersion

No z/OS support

Lower subsystem scalability

Point-in-time and RC availability, but with deployment limitations

Features impact performance

Questionable data consistency and fast recovery availability

Less connections for RC

A failure of a component may cause 50% throughput and connectivity loss.

In-array compute ....

Currently most of the storage Control Units are based on the same technology as the servers; multi core Intel chips. In fact the multi core is used much more effectively is storage CU than in servers. Multicore technology and server virtualization bring some other developments to watch such as the Virtual SAN Appliances (VSAs)* or VNAS** and embedded application on storage control units. The first emulates the server as storage CU, the second using storage CU for applications in Virtual partition or native. Examples :

Remote replication (RecoverPoint on VMAX 40K, 20K), Drive Encryption – EMC, HDS, IBM high-end subsystems, Real-time Compression to IBM Storwize V7000, SVC.

To emulate CU in server partition may be viable solution for SMBs with limited budget.

Future usage of integrated applications:

De-duplication

Server-less, LAN less backup

E-discovery, searches

Analysis

*LeftHand Networks pioneered, evolved as HPs StoreVirtual (VSA), NetApp, OnApp, Nexenta, StorMagic, Mellanox Storage Accelerator VSA product accessed over Ethernet or InfiniBand supports DAS & SAN

** Houston, Texas-based SoftNAS start-up

IBM storage crew: Why bury your BEST kit at the back of the larder?

Power vs Intel

1) POWER7 cluster is more powerful than Intel

2) Economy of scale

3) Why buy from Intel if you can use your own technology

4) EMC use Intel because they don't have own technology. They are designing the hardware to support the Enginuity generation after generation.

Do you really think that a mid range system, with limited processing power can handle same functionality as hi-end storage. The DS8000 has more computing power (Power-PC & ASICs) in their front and back-end adapters than the SVC.

SVC and the V7000 are great products but still not comparable to hi-end.

The problem of IBM in storage is not the products - it is marketing, confidence and sales attitude.

Martin,

Your suggestion to replace the DS8000 with the V7000 in analogy is like installing motorcycle wheels on BMW 7 series.

There is a huge difference in functionality, performance, scalability and availability between a mid-range and high-end storage. FICON interface doesn't convert storage subsystem for mainframe storage.

Some (not all) of the System z features:

* High Performance FICON (zHPF) - zHPF is high-performance data transfer. It has four functions and each function requires cooperation between the System z server and the storage system. Selected zHPF functions are: multitrack (allow reading or writing more than two tracks worth of data by a single transport mode operation), extended distance, format writes, QSAM, BSAM, BPAM, and DB2 list prefetch.

* Parallel Access Volumes (PAV) and HyperPAV are features which allow using multiple devices or aliases to address a single ECKD disk device.

* Performance - I/O priorities Supports "importance" and "achievement" information provided by z/OS Workload Manager

* Performance - DB2 Specialized cache algorithm to optimize DB2 list prefetch operations

* Performance – IMS Enhanced performance for IMS write-ahead data set (WADS)

* Performance - zDAC Supports an optimization to improve performance of z/OS Discovery and AutoConfiguration (zDAC)

* Volume Management Supports dynamic volume expansion for standard (thick) 3390 volumes , Extended Address Volumes (EAV) –supports 3390 volumes up to 1TB capacity

* Multiple Readers for IBM System Storage z/OS Global Mirror - improved performance and fewer disruptions under heavy write load conditions and as a result experience significantly better performance in particular in busy z/OS environments.

* GDPS and HyperSwap support

Since October 2007 IBM has accelerated its development rate, offering enhancements at the fastest pace in the industry. Examples of major enhancements include IBM’s introduction of RAID-6 in August ‘08, high-performance FICON for System z in October ‘08, full-disk encryption and a solid-state drive (SSD) option in February ’09, and thin provisioning July ‘09.

In April 2010 IBM announced and delivered the IBM System Storage Easy Tier which automates data placement within the DS8000 subsystem. This includes the ability for the system to automatically and non-disruptively relocate data (at the extent level) across drive tiers, and the ability to manually relocate full volumes. This was the first sub-LUN automated data movement.

IBM's System Storage DS8870 series is a stable multiplatform high-end storage system . IBM, which ten, fifteen years ago lost some of its high-end disk enterprise storage market share, has managed an impressive come-back.

There are plenty synergies in operation between System z mainframes and DS8870 subsystems in particular in Disaster Recovery and Business Continuity but also with DB2, etc.

The operation GIU, ported from XIV delivers the most user friendly functionalities in the industry.

Full redundancy, non-disruptive upgrades and maintenance, hot-swappable components, pre-emptive soft error detection and online microcode changes ensure high availability and data integrity. The advanced remote data replication techniques enable any scheme of disaster recovery deployments.

The DS8000 is a leader in SPC-1, SPC-2 (non-SSD) performance benchmarks.

LTFS and ugly ducklings

Tape is not disk or Flash and tractor is not a car but each one can be used in optimum way.

The tape is still the lowest $/GB, the lowest energy/GB, supports input/output etc.

LTFS

Self-describing tape format to address data archive requirements

Tape is logically divided “lengthwise” into two partitions

Index partition: File system info, index, metadata (37.5 GB)

Content partition: Contains the files / content bodies (1425 GB)

Implementation of specific software that uses this data format to provide a file system interface to data stored on magnetic tape

Tape format developed by IBM (adapted by LTO consortium)

Defines the organization of data and meta-data on tape - files stored in hierarchical directory structure.

This format makes it possible to implement software that presents a standard file-system view (letter) of the data stored in the tape media

Direct access (drag and drop files) to file content data and file meta-data.

Interchangeability of tapes between systems

Gartner magicians mumble, fling bones, scrawl new disk pentacle

To some extend, but it is NOT product evaluation.

Many users misinterpreting the MQ and using them as product evaluation and procurement help. In reality the MQ are showing more the company performance than product quality. In the past Gartner tried to publish storage evaluation report which was comparing different products and vendor viability. One report was published but it was not continued because of one vendor who argued about almost every criteria (his and other products) which was very time consuming.

BTW, I authored many MQs in the past.

The future of storage

answer and some more

As explained above I didn’t use the right words. Despite the benefits of tape usage such as the lowest price per GB, the lowest energy consumption per GB, portability, fastest development in bit density and new features such as WORM, encryption and Linear Time File System (LTFS), the tape market is constantly shrinking. The major reason is using nearline disks with deduplication as the targets for backup. The idea of virtual tape libraries which combines benefits of disk (as a cache) and tape on the backend didn’t achieve great success outside the mainframe market and is not able to stop the erosion of the tape market.

Currently most of the storage Control Units are based on the same technology as the servers; multi core Intel chips. In fact the multi core is used much more effectively is CU than in servers. Multicore technology and server virtualization bring some other developments to watch such as the Virtual SAN Appliances (VSAs) and embedded application on storage control units. The first emulates the server as storage CU, the second using storage CU for applications. The VSA:

 Emulating shared block-access storage area network (SAN) system on internal DAS storage (running in virtualization partition)

 The server SW acts as SAN array controller software

 A pair of servers with high amount of direct-accessed storage (DAS), accessed by other servers across a network

 LeftHand Networks pioneered, evolved as HPs StoreVirtual (VSA), NetApp, OnApp, Nexenta, StorMagic

 Mellanox Storage Accelerator VSA product accessed over Ethernet or InfiniBand supports DAS & SAN promise better performance

 Saves HBAs, switches, Physical CU price

The usage of embedded applications:

usage of integrated applications:

 Remote replication (RecoverPoint on VMAX 40K, 20K)

 Compression (Real-time Compression to IBM Storwize V7000, SVC)

 Drive Encryption – EMC, HDS, IBM high-end subsystems

Future

 De-duplication

 Server-less, LAN less backup

 E-discovery

 Analysis

Some clarifications

Why … kick a dead horse? Tape is not [a] growing business. The smart tape vendors, like Fujitsu with CentricStor, are not enjoying great success. I bet on “cheap” disks with … mirroring in de-clustered RAID [or] Erasure Coding. [For example] 2.5-inch HDDs (2020 - 12TB, 3.5-inch - 60TB), more platters, all SAS.

Same more details:

The vanilla tape market is shrinking constantly, the only way to slow that is by using smart tapes such as virtual tapes libraries. The current vendors of the VTLs are not selling them in quantities which can stop the erosion of the tape market.

“cheap” disks connected in RAIN (Redundant Array Independed Nodes) structure .

The 2.5-inch will continue as enterprise HDDs not the “cheap” line, the 3.5-inch will continue as “Nearline” capacity disks. Going back to large factors such as the 8.5-inch means more energy, noise, space, not fitting standard rack , super long re-built times, etc.

We were right: EMC's VSPEX will take on FlexPods

IBM's PureSystems propose more flexibility and openes

1) Support more computing platforms; Windows & Linux on x86, AIX & Linux on POWER and System i on POWER

2) Large scalability

3) More applications from hundreds of ISVs

4) 3rd party storage on the virtualization layer of the Strorewize V7000

5) Wide palette of connectivity options

6) Optimized internal network.

Fujitsu Eternus: It's time to settle this thing

some thoughts

Some remarks:

1. VMAX is modular built storage subsystem but who say that this is a benefit to the end user. As opposed to cache centric monolithic control unit, adding modules most probably will increase the response time on some I/O. See my presentation from SNW 2010. Such structure require frequent tunings and redistribution of data.

2. Feedback from one of my customers: VMAX can't scale not disruptively. Updating cache means taking-off the side with the host channels. The customer must provide 2nd path.

3. Fujitsu Eternus series support the same functionality and storage management across the the whole line. VMAX, VNX supports different functionality. The VMAX is controlled by Enginuity, the block part of VNX by FLARE sitting on Windows and the file by DART -3 different remote copy techniques, 3 different flash copy, etc.

4. There is nothing virtual in the Virtual Matrix Interface - it is RapidIO protocol which is competition to InfiniBand but it is very real.

VMware 'to work with just five storage companies'

Till mid 90-ies the storage world for other platforms than mainframe was locked-in, each server vendor supported his own storage. Data General opened this market with the CLARiiON. EMC marketing philosophy was always based on lock-in by features, for example PowerPath was originally created by Conley Corporation was develop as open path management however EMC after acquiring Conley limited the use for Symmetrix only, similar was with the first virtual tape CopyCross which was developed by EMC engineers to support any storage (in fact the initial beta site was on IBM ESS subsystem in Technion Haifa) but later limited to support Symmetrix only.

Is EMC trying to create new lock-in with the help of VMware? It may result in Pyrrhic victory; There are alternatives to x86 virtualization hypervisor (hyper-V, Citrix Xen and Red Hat KVM). If I had today to evaluate virtualization platform for the future I would select open approach.

B+F: EMC bloggers clash with analyst

Logic

To J.T.

Whoever is hiding behind these initials?

First: If you think that the issue is Cereva you just missed the point, this is a result of one of the deviations.

Second: Some facts Cereva Networks was acquired by EMC in June 2002 and not August. Please see below a LinkedIn entry of one of the Engineers:

Staff Software Engineer

EMC

Public Company; 10,001 or more employees; EMC; Computer Hardware industry

July 2002 – May 2005 (2 years 11 months)

Worked for Advanced Development Team and a project coming from the AD Team. Worked in Linux Kernel, Windows XP Kernel. All code written in C, C++.

Principal SW Engineer

Cereva

Privately Held; 51-200 employees; Computer Hardware industry

August 1999 – June 2002 (2 years 11 months)

Storage Startup. Worked in the I/O path. Code written in C on Unix for use in VxWorks platform.

Third: your remark “There's no way that EMC could swallow up IP and put it into their flagship prodcut that fast.”

You are right, it is much more logical that EMC developed a complete new high-end storage subsystem (hardware, software) after the departure of Moshe Yanai in less than two years and EMC acquired Cereva paying 10M$ because they have big heart and wanted to secure the future of 40 Engineers.

You don't know me

Joseph,

If you don't know me why you wrote "Krischer, Hollis and Burke are no different"? I pointed to factual discrepancies and FUD in their block but instead getting viable answers or correction got a coordinated personal attack. There is a big difference between EMC and HDS, IBM, NetApp etc. bloggers- they continuously spread FUD about the competition. On the 25th October I spoke with a customer in Germany who was purchasing storage. He considered EMC and IBM XIV. One of the reasons that IBM won was that EMC pointed all the time to disadvantages in XIV saying very little about their own products. It must be corporation culture.

Anyhow, if you want to know more about me please visit www.joshkrischer.com. I will be also happy to discuss that with you over a phone.

Josh

the full quote

“Josh Krischer demonstrates a knowledge of the server and storage markets that is nothing less than encyclopedic. He is blunt, honest, uncompromising and often irascible - but very rarely wrong!” February 27, 2006

Andy Butler

Vice President and Distinguished Analyst , Gartner Inc. (ex-colleague)