Email: Top Corporate Liability
We revisit the biggest risks IT managers face when it comes to email archiving
October 3, 2007
Yesterday's article on the top mistakes users make when archiving email was incomplete, according to feedback from a couple of readers with strong opinions.
One of these, Mark Diamond, president and CEO of technology consulting firm Contoural Inc., believes we neglected some of the most common and serious errors users make. Here, for the record, is his list:
Failure to perform in-house due diligence. "The single largest mistake is failure to have a policy or consensus before buying a product," Diamond says. Often, he sees IT folk purchase an email archiving product before consulting with their legal, HR, records management, or compliance departments. The result is usually a snarl at the last minute, as various in-house groups struggle to make their issues known, meantime bollixing the IT purchase.
"You need to start engaging a committee from the beginning," Diamond maintains. The best policies for retention of email are set by IT and the legal department, jointly, in his view.
Failure to keep it simple. Diamond thinks many organizations also make policies for email that are too long and complicated. "We see some organizations with 50- to 200-page policies," he says. "It's better to create a simpler policy that's easier to automate and execute." After all, regulators and courts are looking for evidence that a company had an automated process that was transparent and consistent. They don't expect perfection, he notes, but if you had a policy that wasn't followed because it was too intimidating or difficult to understand, that's worse than having no policy at all.
Failure to establish a "litigation hold" policy. A company can create an email archiving system but fail to implement it when it's required that is, when the lawyers come calling. That's what happened, Diamond says, when Intel faced its suit by AMD. Though Intel had a policy for saving email, it failed to notify employees that certain messages would have to be earmarked for legal perusal. The result was that Intel spent over $3 million to go through backup tapes in order to meet court mandates for electronic evidence.
Failure to have reasonable retention periods. Many firms, says Diamond, institute retention periods for data that are too short. "Thirty to sixty days doesn't work," he insists. When there is such a short window in which email is archived, employees may take to "underground archiving" on USB drives, laptops, Gmail, or other untrustworthy places.
Diamond also thinks we erred in suggesting that .pst files be incorporated in email archives. In his view, .pst files represent a risk for any organization because they aren't under the control of IT. It's better not to ingest them into an archive, he asserts.
Another reader, Roger Matus, CEO of InBoxer Inc., a supplier of spam filtering and email compliance products, discourages folk from archiving instant messages – something yesterday's article suggests.
"You state that instant messages should be archived," writes Matus in an email today. "There is a strong argument that says that except for certain regulated industries, you should NEVER archive instant messages. If you look at the Federal Rules of Civil Procedure and other related rulings, the courts insist that if you have an electronic record anywhere in your organization, you must produce it. But, if you do not have a record, it does not need to be produced. Therefore, unless you are already storing IM for another reason – DON’T START. What you are doing is creating new evidence for opposing counsel to subpoena."
Matus, who has a blog of his own relating to this and other topics of email legality, also disagrees with the article's assertion that so-called bacn email is a candidate for elimination from archives.
"You imply that systems should remove 'bacn.' Actually, the FRCP, state open meeting laws, the Freedom of Information Act, and others make no distinction between bacn, spam, and regular mail," Matus writes. "If it is actually delivered to the end user and it is not already deleted from the client, it must be retrieved. You could argue that bacn is never relevant to a court case. But, are you sure? Spam has been used in cases as evidence of a hostile work environment. So, vile spam needed to be produced."
So there you have it: More on the weighty issue of archiving email for compliance and legality. Do you think we missed something? Let us know. You could help to save someone's job. Hit the message board below, call us, or send us a message.
Have a comment on this story? Please click "Discuss" below. If you'd like to contact Byte and Switch's editors directly, send us a message.
Advanced Micro Devices (NYSE: AMD)
Intel Corp.
Alacritech Inc.
ATTO Technology Inc.
Cisco Systems Inc. (Nasdaq: CSCO)
EMC Corp. (NYSE: EMC)
IBM Corp. (NYSE: IBM)
QLogic Corp. (NYSE: QTM)
SGI
Silverback Systems Inc.
Sun Microsystems Inc. (Nasdaq: SUNW)
Tacit Networks Inc. (NYSE: EMC)
Hitachi Data Systems (HDS)
IBM Corp. (NYSE: IBM)
3Leaf Systems Inc.
Pillar Data Systems Inc.
VMware Inc.
Red Hat Inc. (Nasdaq: RHAT)
Virtual Iron Software Inc.
VMware Inc. (NYSE: VMW)
XenSource Inc.
Zeus Technology (Nasdaq: NTAP)
Packeteer Inc. (Nasdaq: PKTR)
Reldata Inc.
Riverbed Technology Inc. (Nasdaq: RVBD)
Tek-Tools Inc.
Holding fast at number five, SANcastle Technologies' sales of its FC/IP translation switch are believed to be creeping up, albeit slowly. EDS is already a customer, and it recently signed a reseller deal with Ethernet switch player Anritsu Corp.
Compaq Computer Corp.(NYSE: CPQ) has also certified SANcastle as an “approved interface vendor” for its Data Replication Management FC-IP Solution. (We don’t actually know what this means, but it sure sounds impressive!)
The official announcement of this approval is not due until January, but you can get the details here, now: ftp://ftp.compaq.com/pub/products/sanworks/techdoc/drm/EK-DRMIP-AA-A01.pdf
For the previous Top 10 article about this company, see:Lower The Drawbridge! InterSAN Inc.
The second new entry on our list pushes out 3PARdata Inc., taking the number six spot. InterSAN is working on software to manage storage systems in networked environments. MerrillLynch & Co. Inc., IDC, and the The Yankee Group all think that this will be one of the hottest areas in storage networking next year. For once, we agree.
InterSAN offers automated storage provisioning software that tackles a major part of the storage management issue. It lets users automate storage provisioning, measure service-level agreements, and create and monitor policies across heterogeneous storage networking components (see InterSAN Attracts Major InterEST).
Basically, this is spicy hot stuff! EMC Corp. (NYSE: EMC) is attempting to do similar things via its AutoIS software initiative (basically, an invitation to other storage vendors to share upper-level management APIs in order to manage storage across the different systems). The problem EMC faces is getting its competition, which it has so vehemently disregarded in the past, to now go ahead and work with it (see EMC Goes Soft).
While EMC fights its political battles, InterSAN has a great opportunity to take a leadership stance in this market.
Alacritech Inc. is the last new entry on this list, making its debut at number seven. This company is keeping everybody on their TOEs with a gigabit Ethernet SAN NIC accelerator card, also known as a... erm, TOE (or TCP/IP offload engine, to be really flash).
As well as consolidating NAS file data and SAN block data over an IP network, it also comes with an IP stack, offloading IP processing tasks from the server and improving overall network performance (see Alacritech Demos Accelerator). This is a big deal, because right now IP overhead is the biggest roadblock to widespread deployment of IP in storage nets (see Sun's Down on Storage Over IP).
Then there’s the little matter of who happens to be running this company: It’s only Larry Boucher -- the inventor of SCSI, the founder of AdaptecInc. (Nasdaq: ADPT, and now a billion dollar company), and, prior to that, a major cheese in the storage division of IBM Corp. (NYSE: IBM) for 12 years.
Alacritech faces competition, though. Adaptec bought Platys, which is working on the same sort of product; and several other startups -- including Trebia Networks Inc. and Silverback Systems Inc. -- are hot on its heels. But right now, Alacritech looks to have the lead and is already shipping product.
is still undeniably hot, but that hasn’t stopped it dropping six places in this listing.
Partly this is because the industry expectations for BlueArc are so high. It made a huge splash when it launched, received buckets of funding, and has a tip-top management team. But sales, since it started shipping in July, have been slow.
BlueArc has shipped five of its Si7500 storage servers to Lawrence Livermore National Laboratory (LLNL) to be used in conjunction with LLNL's accelerated strategic computing initiative (ASCI). BlueArc has confirmed that LLNL did pay for these boxes.
However, other sales have gone to strategic partners, such as Extreme Networks Inc.(Nasdaq: EXTR) and Juniper Networks Inc. (Nasdaq: JNPR) that are demonstrating the equipment in their respective labs as a favor to BlueArc.
A company spokesperson told Byte and Switch that BlueArc is under increasing pressure from Network Appliance Inc. (Nasdaq: NTAP), which is shipping boxes for free, or almost free, to counter the competition.
What’s needed for BlueArc to move back up the list? A few big customer wins.
For the previous Top 10 article about this company, see:Big BlueArc
On September 19, we wrote up 3PARdata Inc.’s $100 million round of funding, giving it the title 3PARdata is Ready. Did we speak too soon?
3ParData has more money than most of the startups on the Top 10 put together, and an impressive management team to boot. But there’s no sign of a product yet, not even a whisper of what color it might be!
It’s time for 3PAR to put up or fall off.
For the previous Top 10 articles about this company, see:EMC Hammer and
Hmmm. Where to start? What a year for Cereva Networks Inc. Here’s a snapshot of the stories we’ve run. They should give you a good indication of why it has dropped down two spots to number ten.
Cereva Details Storage Switch
Cereva: Stalled, Not Stopped
Cereva Closes In On Round Three
Cereva Skids to a Halt
Cereva's Back on Track
Cereva Exec Churn Continues
Cereva Cuts Twenty MoreTime will tell whether the new management team at Cereva is better than the old. Cereva is head-to-head with 3PARdata and, like 3PAR, details of its product are long overdue. It’s time for the company to talk about its technology, or we’ll be shipping it out (and off) our top 10 list.
For the previous Top 10 articles about this company, see:
Viva Cereva! and
IN THE BIT BUCKET
Off the list and into the bit bucket – plip! — go TrueSAN Networks Inc., Troika Networks, and Zambeel Inc.
Yotta Yotta Inc. abides, largely because of the Yotta Yotta Song.
TrueSAN was first to market with a patched-together SAN-in-a-can (a large storage array and switch in one box a la 3PAR and Cereva) but has still not announced any customers. Troika launched its proprietary server clustering technology, but since then there’s been no word from this company, which isn’t a good sign.
And Zambeel got a ton of money from Kleiner Perkins Caufield & Byers. The startup is also expanding into the lucrative Japanese and Korean markets. (See Riverbed Opens in Japan and Riverbed Launches in Korea.)
Riverbed is also gaining plenty of partner traction, opening up a broader channel to market than its rivals. Riverbed recently signed a deal for McData Corp. (Nasdaq: MCDTA) to OEM Steelhead appliances. (See McData Hits Remote Control and McData Chases Remote Users.) This partnership followed a major OEM deal with Hewlett-Packard to license Riverbed's core software on HP's StorageWorks Enterprise File Service (EFS) WAN Accelerator. (See Riverbed Licenses Tech to HP and HP in Deal With Riverbed, Sources Say.)
But Riverbed does not want to be seen as acquisition fodder. The firm apparently isn't for sale or looking for investors. CEO Jerry Kennelly has been sending signals through the financial community that he wants to take Riverbed public this year. Kennelly told Byte and Switch earlier this year he expects to be profitable by the end of 2006. (See McData & Riverbed: A Rumored Pair.)
Mellanox Technologies Ltd. (Nasdaq: MLNX) continues to climb the storage market on the strength of InfiniBand silicon.
The seven-year-old firm abandoned its end-user business entirely last year in order to focus on delivering parts to suppliers of InfiniBand gear, including Cisco Systems Inc. (Nasdaq: CSCO), Dell Inc. (Nasdaq: DELL), Engenio Information Technologies Inc. , Hewlett-Packard, IBM Corp. (NYSE: IBM), Isilon, SilverStorm Technologies Inc. , and Voltaire Inc.
The strategy seems to be working. VP of product marketing Thad Omura claims the OEM customer count has doubled since August 2005, from 100 to over 200. Employee headcount has gone from 140 to 160.
“Year on year, we’ve had 100 to 150 percent annual revenues for four consecutive years,” Omura boasts.
The question is, Can the momentum continue? For now, InfiniBand appears to be sustaining popularity, particularly in data centers where clustering is used as an alternative to supercomputers.
According to Omura, the largest cluster in 2005 – 4,000-plus nodes at Sandia National Laboratories – has been quickly superseded in at least two or three sites.
But there’s a gorilla looming on the InfiniBand horizon: 10-Gbit/s Ethernet. Smaller players are starting to deliver systems, and chipmakers are signing up deals. (See 10-Gig IP SANs Hit Bleeding Edge, NetXen Singles Out 10-Gig, and Force 10 Fires Up Low Latency Switch.) Given that most supercomputer sites still use Ethernet instead of InfiniBand, does this worry Mellanox?
“We always see a threat from 10-Gbit/s Ethernet,” Omura quips. But, he insists, 10-Gbit/s Ethernet is not close to catching up with InfiniBand on price/performance. Latencies remain higher than InfiniBand’s, and costs of 10-Gbit/s Ethernet NICs remain above $2,000 per port. InfiniBand, meanwhile, is already running at 20-Gbit/s and above and delivering the goods for less than $200 per card. (See Mellanox Doubles InfiniBand.)
InfiniBand is way ahead of its chief rival, Omura says. What remains to be seen is whether Mellanox can build itself up fast enough to ensure its future even when and if 10-Gbit/s Ethernet starts ruling the SAN. One approach may be to go public, a move that Mellanox is rumored to be considering. Omura won’t comment, except to say, “We certainly are driving toward that.”
Sepaton Inc. earns its position on the Top 10 list thanks to strong customer traction in the burgeoning virtual tape library (VTL) software market, as more and more users seek alternatives to tape that take advantage of low-cost SATA drives.
The Southborough, Mass.-based firm, which was known as SANgate Systems until late 2003, certainly seems to be heading in the right direction at a time when users are waking up to the benefits of VTL software. (See Sepaton's Got Virtual Tape, and SANgate Tries Again.)
Earlier this year, for example, Sepaton announced that it had reached the 150 customer mark, with more than a smattering of big-name clients. These include AAA, China Construction Bank, Citistreet, Foxwoods Casino, Harris Corp., KeySpan, and telecom giant MCI Inc. (Nasdaq: MCIP). (See Sepaton Claims 150 New Customers .)
One Sepaton customer told Byte and Switch that, compared to other vendors in this space, the startup benefits from faster speeds for restoring data. Michael Grillo, principal IT engineer for Foxwoods Casino, says Sepaton won a bakeoff over Neartek Inc. 's Virtual Storage Engine and EMC's Clariion Disk Library (CDL) last year – mostly because it produced faster restores.
"Speed was the heavy hitter in making our decision," Grillo says. He also liked that Sepaton scaled easily, as lack of scaleability is a frequent complaint with VTL products. "I initially purchased 40 Tbytes and bought another 20 because of growth four months later. It came in, I plugged it in, did one boot, and I had an additional 20 Tbytes of disk."
Initial signs are that Sepaton's OEM partnership with HP is also starting to bear fruit. Sepaton also has added remote-site backup and extended its VTL capacity. (See Sepaton Goes Standard on VTL , HP Upgrade Features OEM Crowd, HP Integrates Sepaton, Sepaton Enhances VTL, and Sepaton Adds Remote Rep.)
On the financial side, the startup is going from strength to strength, recently claiming 300 percent annual revenue growth and, in November, racking up $15 million in VC funding to support its product development. (See 2005: The B&S Report Card and Sepaton VCs Raise Their Bid.) That round brought Sepaton's total funding to $68.6 million.
CEO Mike Worhach won’t rule out going back for more funding, but he projects Sepaton to reach breakeven by the second quarter of this year.
But Sepaton is not the only private company making waves in this part of the market. The startup faces increasing competition from an expanding list of VTL competitors as well as other disk-based backup systems presenting alternatives to tape. Rival startups Copan Systems Inc. , Diligent Technologies Corp. , MaXXan Systems Inc. , and Neartek, for example, remain in the game. (See Copan Cops $17.5M.) Then there are compression vendors such as Avamar Technologies Inc. and Data Domain Inc. (Nasdaq: DDUP), which claim to reduce the disk space needed to back up data. Watch this space.
If you’ve been shopping for backup that goes a little faster, works a little harder, or makes more efficient use of disk resources, chances are pretty good you’ve run into a vendor pushing data protection.
While there are lots of companies in this space, we’re singling out Data Domain Inc. (Nasdaq: DDUP) for its technology, the traction it’s gained with customers, and an August 2005 round of funding that brings its capitalization to $41 million. (See VCs Add $15M More to Data Domain.)
This is not to say that Data Domain is head and shoulders above its many competitors, which include Asigra, Avamar Technologies, and Diligent Technologies Corp. If Data Domain has any edge, it’s a slight one, and the momentum in this niche may very well shift six months from now.
The term “data protection” gets liberally interpreted to include just about any company that ever set foot in the backup space, but as this slice of vendors (re-)defines it, data protection offers more than regular weekly backups and daily incrementals. First, these vendors may also offer compression, either at the application server where the data originates or on the disk end where the stored data resides. Further, they claim to handle de-duplication that preserves data integrity. And they say they can provide backup more cheaply than tape or virtual tape vendors can.
Data Domain’s portfolio is small: The DD460g, a gateway server that compresses data for storage on customer disk arrays, is the newest product. (See A Storage App Without the Storage.) The vendor considers the DD400, an appliance with a built-in disk, its flagship product and says it will discontinue the DD200. The vendor boasts customers as varied as Children’s Hospital Boston, the accounting firm of Mohler, Nixon & Williams, and Quicken Loans Inc.
Data Domain products only became commercially available at the end of 2003, but there are now 800 systems in the field and some 250 customers, according to Brian Biles, vice president of marketing. “We’ve grown 600 percent year over year,” he contends, noting one third of revenue in 2005's fourth quarter came from non-U.S. customers, with another third from existing customers buying upgrades or extensions.
“Data Domain’s claim to fame, without question, is the compressability of data and, at the same time, the non-disruptive quality of that capability,” notes Arun Taneja, founder of storage consultancy the Taneja Group . He notes that Avamar compresses at the application server to minimize the amount of data that’s then transferred across a local or WAN link for storage; the downside here is it uses CPU cycles from the app server. Data Domain handles compression on the storage end of the equation, and as such, is less disruptive to the way most storage managers think of, or execute, backup.
“Data Domain doesn’t ask the customer to change procedures as Avamar does,” Taneja explains. “Any time you ask them to change, you get pushback, so Data Domain has had a little easier liftoff.”
Still, he credits both vendors with evangelizing the whole compression phenomenon, which just about overnight has become a “must-have” for larger, more established storage vendors. Witness the premium paid for startup Rocksoft in mid-March, suggesting there was a bidding war that ended $63 million later, according to Taneja. (See ADIC in De-Dupe Deal.)
Biles doesn’t flinch when pressed about whether his company is being groomed for acquisition. “We’re early to market with a core platform technology that a lot of people will realize they need somewhere down the line,” he says. The DD line tripled capacity and doubled throughput from first generation to second, and that’s a feat the company will repeat a few times in the next couple years, Biles claims.
“The only thing we think about is growing to be a large, independent company. There are various eventualities, but staying independent gives us better leverage against whatever else comes up,” Biles says.
Of the startups currently vying for dominance in continuous data protection (CDP), Revivio Inc. stands out for its ability to command enterprise respect.
While CDP is still relatively immature, there is growing momentum behind the technology, which captures all the changes made to data. The idea is that this allows users to restore any good version of a document or application in case of a system or media failure.
Unlike virtual tape libary (VTL) backups, which usually activate just once a day, CDP technology, as its name suggests, works on a continuous basis, monitoring write-and-log I/Os with time-stamps to enable recovery from any point in time.
Revivio, unlike many of the other CDP vendors, handles data at a block – as opposed to file – level, enabling quicker backups and broader application support. Another startup, Mendocino Software , also handles data at the block level, but that vendor's OEM deals with EMC and HP are only just getting off the ground. (See EMC Pulls Forward With Backup and HP Picks Mendocino .)
Other private companies playing in this space include CA XOsoft and Zetta Systems Inc. , but Revivio gained the most VC-backing of the three last year, completing a $25 million Series C. This brought the startup's total funding since its October 2001 inception to $55 million. (See Revivio Revs Up With $25M, NAS-tronomical Year in Funding, and Revivio Revs Up .)
Things are also progressing nicely on the customer front. Revivio has clinched deals with the University of New Mexico, Forbes.com, Baptist Memorial Healthcare, SAS Institute Inc., and PrairiePackaging. (See Firm Rolls With Revivio, Baptist Memorial Healthcare, Hospital Prescribes VTL, CDP, Healthcare Prescribes Storage, and Revivio Turns SAS Onto CDP.)
The Lexington, Mass.-based firm has also been racking up partners at a clip, announcing deals with Intel Corp. (Nasdaq: INTC), VeriStor Systems Inc. , Data Protection Associates, and Networks Plus Technology Group, among others. (See Revivio Shifts into High Gear , SANZ to Resell Revivio, VeriStor Offers Revivio CDP, and Revivio Joins Intel Community.)
The next few months, however, will be crucial for Revivio, as big-name vendors like EMC and HP crank up their own CDP strategies. But at least the startup is planning for tough times, recently appointing a new CEO, Terry Leahy. (See Revivio Names CEO , Microsoft Backs Up on CDP Claim, IBM Hops CDP Bus, and 2005 Top Ten: '06 Storage Predictions.)
In this climate, Revivio might just be ripe for picking by one of its larger competitors. Network Appliance has already flashed its cash for Alacritus, and earlier this month, Atempo Inc. swallowed up Storactive for an undisclosed fee. There's a chance Revivio could well be next in line to get snapped up. (See NetApp Annexes Alacritus, NetApp Readies Virtual Tape, and Atempo Swallows Storactive.)
But who knows what things will look like next year? At the moment, user feedback suggests that Revivio is winning business thanks to its strong customer base and the fact that the vendor has a clearly identifiable product. This situation, however, could change significantly a year from now when Mendocino's OEM deals are in full swing.
Onaro Inc. is in Phase Two as far as its products and business strategies go.
The first phase – under founder Shai Scharf – consisted of bringing out an application that handled change management in storage and winning acceptance for that new concept. (See Shai Scharf, CEO & Co-Founder, Onaro and Onaro Ships Change Manager.) Now Onaro has a new CEO and is looking to demonstrate that its SANscreen platform can do more than just manage change in SANs.
Doug McNary, named Onaro CEO last October, has a strong sales background. He headed sales at bandwidth management software company Motive and Web publishing firm Trellix and also held sales executive positions at Tivoli Systems, Technology Concepts, Wang, and General Electric.
McNary says Onaro has around 40 customers, with many of them large installments, such as State Street Global Advisors and Priority Health . McNary expects to at least double that by widening the product line and expanding into Europe and Asia.
Onaro upgraded its core product, recast its business continuity module, and added a reporting tool earlier this month. (See Onaro One-Ups SRM.) McNary expects to launch another product near the end of the year. He says the new products are built around treating storage as a service.
McNary is also pursuing partnerships that could take Onaro beyond SAN management into NAS and even networking.
"Our enabling technology is designed to address many aspects of the IT discipline," he says. "It's a modeling engine with advanced change management capabilities. For now, we’re focused on the storage space. Anything we would do beyond that would be in deep partnership with somebody."
EMC – with its Smarts software – and other large storage vendors have designs on going the same way as Onaro. (See EMC Smartens Its NAS.) The key for the 40-person startup is whether the large guys see Onaro as competition or potential partner. Will the established storage vendors buy change management capabilities from Onaro or develop them on their own?
"There are a number of major storage vendors that recognize that this is an area they need to address," McNary says. "There are various ways of addressing that. One is working with us."
In every iteration of our Top Ten, there are inevitably a few companies that fall off the list, sometimes softly, sometimes with a loud, embarrassing thunk. But note, please, that landing in the Bit Bucket does not necessarily mean we think companies aren't doing well.
Still, this list isn’t about well. It’s about outstanding.
So, with the hautiness of a Simon Cowell, we proceed to name our three Bucketeers:
GlassHouse Technologies Inc. continues to add customers and services. It’s boosting its sub-business in taking over support duties for some storage suppliers. Yawn. Wake us up when this firm does something to change the nature of networked storage – or better yet, when it goes public.
After a running start in the network file virtualization market, Acopia Networks Inc. has settled into a slow, steady jog. Indeed, the heat’s gone out of this race for the time being. Rivals Rainfinity and NuView have been bought by EMC and Brocade, respectively, leaving Acopia and NeoPath Networks as the chief standalone offerings in the space. (See EMC to Buy Rainfinity and Brocade Bags NuView.) Did Acopia miss the boat? We think not. Still, it will be hard work competing against major players that opted for someone else’s technology.
What can we say about Asigra Inc.
Sun Microsystems Inc. (Nasdaq: SUNW)
Taneja Group
Trusted Data Corp.
Woven Systems Inc.
XenSource Inc.
Zetera Corp.
Read more about:
2007You May Also Like