In this episode of #CaseoftheWeek we analyze whether a party that moves data to a new system during the preservation period is liable for spoliation under Rule 37(e). The case we will be discussing is Oracle USA, Inc. v. Rimini Street, Inc., 2020 WL 9209714 (D. Nev. 2020), September 21, 2020.
Watch Next Episode Live
Catch the Replay
You’ll be able to read episode summaries and watch the recorded show right here on the eDiscovery Assistant blog or access the recorded versions from the ACEDS social media channels.
Transcription of #CaseoftheWeek Episode 22
Good morning and welcome to our #CaseoftheWeek for April 27th , 2021. It’s the last week of April. Hard to believe how quickly 2021 is flying. I hope it’s going well for you and your families.
I’m Kelly Twigger, the CEO of eDiscovery Assistant and the principal at ESI Attorneys and each week, in partnership with ACEDS, we bring to you a #CaseoftheWeek selected to point out some practical applications for you, your clients, and how you can improve your practice as it relates to eDiscovery.
This week, we’re going to cover a decision from the 𝐎𝐫𝐚𝐜𝐥𝐞 𝐔𝐒𝐀, 𝐈𝐧𝐜. 𝐯. 𝐑𝐢𝐦𝐢𝐧𝐢 𝐒𝐭𝐫𝐞𝐞𝐭, 𝐈𝐧𝐜 matter, and you’ll find a link to the decision from September 21st, 2020 in the comments section on the platform that you’re watching us, no matter which platform we should all be there.
You’ll also find a link to our 2020 Annual Case Law Report if that’s of interest for you to download. That was created in conjunction with Doug Austin and eDiscovery Today and includes links to the key cases from 2020, as well as some great statistical overviews about what happened with the case law last year. We look forward to putting together a midyear report for you as well in 2021. Lots to report on.
OK, let’s dive into our #CaseoftheWeek. As I mentioned, we’re covering one of the most recent decisions in the Oracle USA vs. Rimini Street decision in this case. This is a case that has a very storied history. It has been going on since January of 2010, has been all the way up to the United States Supreme Court and back down to the trial court. At the heart of the legal fight the question is whether Rimini Street can provide the third party support services for Oracle’s PeopleSoft software without violating Oracle’s intellectual property rights.
Essentially, it’s an enormous decision from a perspective of third parties like Rimini Street providing services related to manufacturer’s software. The PeopleSoft software, like many of Oracle’s products, is enterprise based software where you don’t actually purchase the software, what you purchase is a license to the software and the ability to customize it. Third party providers often come in and provide services associated with that software. We’ve seen it a lot in the Microsoft world. We see it a lot with Office 365 services in the ediscovery space. It’s a similar kind of arrangement, very complex, and the question here in this litigation is what was the scope of what Rimini Street was permitted to do under Oracle’s licensing agreement?
The argument that Oracle made is that Rimini was essentially providing cut rate support for its customers using the PeopleSoft application through unlicensed copies of the software, and that it was essentially using one of the customer’s licensed software to support other customers.
In August of 2018, that’s eight years after the case is filed, and there there are other decisions in eDiscovery Assistant than the decision that we’re covering today that you can look at, we’ll talk a little bit about. But in August 2018, Oracle won a permanent injunction against Rimini Street and was awarded $90 millio from Rimini Street as a result of their violations.
In April of 2019, Oracle asked a federal court to allow a reexamination of Rimini Street’s business to make sure that it’s following the injunction. Rimini Street is now pushing for a jury trial in that case. As I mentioned, this has been up and down to the United States Supreme Court on an issue of what kinds of fees can be awarded. A lot has happened in this case and it’s something that if you google Oracle versus Rimini Street, you’re going to find a lot of business articles about the implications of the intellectual property rights and the contractual relationship between the parties and how that will affect services business for enterprise software.
The decision that we’re looking at here, as I mentioned, is dated September 21, 2020. This is a decision from United States Magistrate Judge Cam Ferenbach out of the District Court of Nevada. This particular decision is a motion for sanctions for spoliation under Rule 37(e) and it’s based on a very specific loss of data. That’s really what I want to focus on. The larger picture of Oracle versus Rimini Street is really important to put this in context, because we’re talking about the actual consulting work that Rimini Street was doing related to leveraging licenses of Oracle software for use at customers.
In essence, when Rimini Street is contracted to come in and do work, the customer would already have a license or should already have a license to the software. Then Rimini Street would leverage that license to be able to help them customize their software and perform the work. Lots of tech involvement here, lots of understanding of how the data transfers would occur in order to be able to understand the ESI related issues in discovery in this case.
On this particular motion for sanctions, Oracle is alleging that during the permanent injunction period Rimini used an automated framework called TransferFiles to copy PeopleSoft files from Rimini’s system to customer associated environments.
Essentially, Rimini does work on its own environment and then copies files back to customer associated environments. They’re arguing that when the work is done on the Rimini computers, they’re taking a file and using TransferFiles to copy that file over to the customers associated environment. The question here is whether the the copy of the file that’s made or what we’ll call the clipboard version of the file, because that’s how the court refers to it, is needed to be preserved for purposes of discovery. Rimini makes a copy of the file. It lives somewhere for a little bit. It’s copied over to the customer’s environment and then that clip file is deleted by the TransferFiles software. A file continues to live on Rimini and a file continues to live on the customer’s environment, but the copied version that was in that clipboard is deleted by the TransferFiles software. Oracle is arguing that that file that was kept on that clipboard should have been preserved and produced in discovery.
A little bit of background to understand the basis for this. Rimini had been using the TransferFiles program to move customer associated data since 2013 and Oracle knew about that. From that time period, from 2013 until 2019, which was the first time that Oracle notified Rimini that it should be preserving that copy data, Rimini was not preserving it. Immediately upon receiving notice from Oracle that it believed that Rimini should be preserving those copies in the TransferFiles protocol, Rimini altered the TransferFiles to preserve those copies and then started producing them in discovery as of October of 2019.
There’s not a lot of time discussed in terms of actual date. We usually talk about on the #CaseoftheWeek that following the timeline of exactly what happens in the case is really crucial. At some point the parties raise this issue. It couldn’t have been too long after October of 2019 because we’re in 2021 at this point and we got a decision on this case in September of 2020. At some point after Oracle raised the decision to Rimini in October of 2019 and Rimini started producing the files. Eight months later, Oracle brought a motion for spoliation to the court, alleging that Rimini should have kept all of these copies of files.
The issue really comes down to one, whether Oracle waited too long to file the motion for sanctions under Rule 37(e) claiming failure to preserve, and two, whether the information that it claims should have been preserved was, in fact data that was required to be preserved under Rule 37.
Rimini looked at the issue and essentially likened it to the clipboard that I mentioned earlier, the notion that in a word processing tool, when you copy data, you paste it to a clipboard and then you paste it to the next place where we want to move the data after that, the data is not retained in the clipboard. Oracle’s flip of that was to say, no, it’s like deleting attachments to an email. You’re essentially spoliating data that you can’t get back.
Rimini’s response was we have the copy of the file that we still have, that we copied using TransferFiles sent to the client, and we have the copy from the client. You could go to the client and get the copy and we’ll help you go to the client and get the copy, and we’ll even agree that you can reopen discovery to go to the clients and we’ll help you get those copy the files so that you can see the difference between what it is that we copy from our servers that we have produced to you in discovery and what it was that the clients were given.
Oracle did not want to do the third party discovery. They did not do the third party discovery during the time that discovery was open in the case, and they did not want to do the third party discovery in advance of this motion. The court even said, do you want to reopen discovery? Oracle said, no, it’s going to lead to too much more delay, another six months to be able to try and get this information from third parties.
Rimini Street tried to even short circuit that and say that they would go to the clients and attempt to help get those files because it was only a file. In some cases it was more than one file, but it should be a relatively clean process.
There’s never an analysis in this particular decision of what difference there is between the file that lives on Rimini Street servers that was provided in discovery and the client based file. That analysis is really where I see the gap is here that causes Oracle’s motion to fail for spoliation.
The court kind of looked at the arguments of the copy and paste process and whether or not there was this was actually transitory information and held a video status hearing. At the video status hearing one thing that the court notes that I think it’s important for us to consider here is that Rimini Street provided a demonstrative exhibit outlining its factual position on the issues. The court doesn’t really go into exactly what that demonstrative exhibit included, but it did say that Oracle didn’t really respond to it during the hearing, responded to it after the hearing, and Rimini Street then objected to that response.
It feels like there’s evidence that Rimini Street was able to put into the hearing that Oracle did not respond to or was not able to respond to, maybe because of the video hearing situation. You don’t really know how things were exchanged, but the fact that the court notes it here seems like it might have been really important.
A few things that the court noted during its status hearing that Rimini had offered to reopen the discovery to allow limited discovery on this issue, and Oracle said no, Rimini had offered, as I mentioned, to facilitate the process of getting these files from the clients, and Oracle said no. Then we also had the demonstrative exhibit.
Then, in terms of the analysis, which the court did on the briefs, they looked really at the factors for spoliation under Rule 37(e). One, did the party fail to preserve ESI that should have been preserved in anticipation or conduct of litigation? Was the information loss because the party failed to take reasonable steps to preserve it? Can the side be restored or replaced, and were the plaintiff’s prejudiced? Those are factors that we’re looking at.
One thing that I think is really important here is that the court cited to the Sedona Principles and noted that the parties are not required to preserve each instance of relevant ESI, and this is an important quote from the Sedona Principles.
ESI is maintained in a variety of formats, locations and structures. Many copies of the same ESI may exist in active storage, backup or archives. Computer systems manage data dynamically, meaning that the ESI is constantly being cached, rewritten, rewritten, moved and copied. For example, a word processing program usually will say the backup copy of an open document into a temporary file every few minutes, overwriting the previous backup copy. In this context, imposing an absolute requirement to preserve all ESI would require shutting down computer systems and making copies of data on each fixed disk drive, as well as other media that are normally used by the system.
Now that quote is especially interesting because it was written a number of years ago before we started having cloud based systems that captured every version of data. We kind of have to take that quote and juxtapose it to the case that we talked about two weeks ago (Nichols v. Noom) where we talked about hyperlinked documents within the context of emails and we know now that those backup copies are not overwritten. Instead, each version of them is preserved. We take that quote in the context in which it was written and the technology that existed at the time. That being said, as applied to this particular situation, it’s applicable, because this isn’t a situation where the TransferFiles program was creating different versions that could be maintained or it was, but TransferFiles was not set up to capture or to keep those duplicate copies until Oracle told Rimini Street that it thought it needed to preserve them.
The Court has to really look at should Rimini Street have thought these needed to be preserved in the first place? The court then looks at a couple of different questions on case law. One, it says that sanctions are not available when parties can restore or replace the ESI through other means and, two that the courts have declined to award sanctions where the moving party has failed to take discovery to try and replace the deleted ESI.
What are we thinking about here? We’re thinking about the fact that Oracle refused to go to the third party clients and get discovery of their individual files. Instead of going and doing that they argued delay and they argued that the clients will have used the files and altered the ESI such that a comparison would not be appropriate and it would not be in the original format.
From a perspective of someone who collects data, I find that argument to be problematic because generally speaking, when we capture files, we tend to keep a preserved copy of the actual file that’s brought in as a backup and then we would use another active copy of the file to be able to load into whatever program you’re using that then you would add new data to. It’s very likely that had you gone to the client, you might have had at least one or two clients might have had that actual preserved file that was exactly the way they would have received it from the transfer file system.
Here, the court also notes that Rule 37 does not contain an express time limit for filing a motion for sanctions, but it does note that there are factors to assess the timeliness of motions. Interestingly, it doesn’t really go through them, but it does say in not so many words that it felt like Oracle’s waiting for eight months after the last time it talked to Rimini Street about this issue, to file the motion for sanctions was not very timely.
In essence, the court adopted Rimini’s analogy that the files were merely transitory and that intermediate files as a credible and good faith explanation regarding how their engineers use TransferFiles to transmit files to a client and how it creates a new copy on the client system while retaining the original file.
Since the court found that these were transitory files there was no duty to preserve each instance. It also found that the data was also duplicative of the files on both Rimini’s system and the client system. Again, if Oracle had been able to show some sort of fundamental difference between the files that existed on the client system, the files that were produced by Rimini in discovery, and how that would impact the theories of liability for trial, that would have been much more telling point for the court to be able to look at on a spoliation motion.
The court also cited the delay in not taking the discovery and in not filing the motion for eight months. They also found that there was no evidence of bad faith by Rimini because it was open and honest with Oracle and that Oracle knew about the use of the TransferFiles system since 2013.
What are our takeaways here?
It’s always easy for Monday morning quarterbacking, right? Which is in some instances is what we do on the #CaseoftheWeek, but our goal is really to look at these cases from a practical take away perspective and give you instances to evaluate what are the things that you need to be considering on a day-to-day basis when litigation arises as it relates to ESI. Here, if Oracle knew about the use of TransferFiles from 2013 onward, there’s a huge disconnect between not realizing that the protocol made a copy of the data that they felt like should have been preserved from 2013 all the way to 2019.
Now, as I’ve already explained, there’s a huge procedural history in this case, very little of which had to do with ESI related issues. There are some other discovery motions, but they’re not on this issue. This happens all the time. We realize after discovery is closed or at the end of discovery, that we have open issues that we have not followed up on, and what’s the process that needs to be followed there?
We talked about this a bit in a couple of the previous cases of the week. You need to go to the court and explain what’s happened. You need to say, Judge, we’ve identified this issue and we need time to be able to dig into it a little bit before discovery closes. We need an extra two months just on this particular related issue or we need to work with Rimini Street to identify X number of clients to be able to see if we can get those files. We’ll report back to you in six weeks on the progress of that. Come to the court with a plan on how to resolve the issue. Instead, don’t wait until discovery closes and then make a motion eight months later without being willing to go back down the discovery path. There’s a big difference here in approach and the reality is, is that as litigators, we have so many responsibilities and these ESI issues literally just add a huge chunk of information onto what we’re supposed to be responsible for and identify. And if you’re not a particularly technologically savvy litigator, you may not identify these issues until much later on, or you may have an expert or some third party who helps you identify them.
Either way, you’ve got to try and deal with them during the discovery process or bring the bring the issue to the attention of the court prior to the close of discovery. Otherwise, you’re going to have a challenge trying to make arguments for yourself afterwards. And that’s what happened to Oracle here.
Another takeaway, you really got to get into the technology here and understand what’s happening, I mentioned it a couple of times. Two things. One, the time period over which Oracle didn’t recognize that this was an issue. Seven years into the case or 10 years into the case, we’re just now recognizing that it’s an issue. Two, Oracle didn’t have the data to be able to show what the difference was between the files that actually prejudiced them as it related to the theories of liability in the case. Two significant issues. You’ve got to get into the data. You’ve got to get into the technology, be able to legitimately show the court what the problem is.
It is really key here, and I think that it’s very important to consider that you are constantly making a record for appeal when you’re at the trial court level. Because discovery issues are never going to be dispositive of a case. If a discovery issue ends up being what turns the tide at trial, you need to have that issue preserved for purposes of appeal. Having the information in the record as to what the difference was between these two files and how it might have impacted the theories of liability for Oracle will be difficult for it to preserve on appeal at this point.
We talked a little bit about how you have to stay on top of the ESI related issues in litigation, and so one of the things that we talk about, one of our themes here is always plan, plan, plan. Have somebody whose job it is to be focused on those ESI related issues, have your eDiscovery department, have your litigation support manager, have someone who understands the litigation strategy here that can help you identify some of these issues. It’s highly likely that someone finally looked at the TransferFiles protocol, realized that these weren’t being preserved, had Rimini Street preserve them, recognize there may be a difference between what Rimini Street had and what the clients had, but didn’t have the ability to go back and get the client files to be able to make a code by code comparison to what existed there.
Then you have to make strategic decisions about do you want to delay the case another six months? It’s already been going on for 10 years. We’re talking about copious expenditures of money. There’s permanent injunction in place. Is that really going to be the end all be all? There are always litigation strategy decisions that are being made. Have someone be focused on those ESI related decisions so that you can be on top of them as you move forward through discovery.
One last take away, one of the biggest challenges that we have as lawyers is giving ourselves the time to stop and think completely about the issues in a case, and that is your job. That is your job as a lawyer. That is what you are paid to do by your clients, is to think strategically about all of the issues and how they all come together in the case for the benefit of the client. That may be to tell the client that they have a great case, to tell the client that they have places in the case where they have potential liability, where they’re unlikely to be able to recover all of those things, but that part where you think and spend your time coming up with an appropriate strategy based on what you know, that’s going to be the greatest time that you can spend for the client. Take that time to think through exactly what it is that you need and how you’re going to need it.
In this particular case, on this particular issue, it feels like had more work and more information they had come to the court with that they would have had a better shot on a motion for sanctions if there was some difference between the files. Based on what we see in the decision, and the types of transitory files that we’re talking about, without being able to point to what would have been different in the conversion from the Rimini Street file to the client file There doesn’t seem to be much of a basis for motion for sanctions here.
That is our #CaseoftheWeek for this week. Thanks so much for joining me. I’ll be back next week with another decision from eDiscovery Assistant. If you’re an ACEDS member and interested in using eDiscovery Assistant, there’s a discount available to current ACEDS members and a trial for folks who are taking the ACEDS exam. If you’re interested in either of those, please drop us a line of ACEDS@eDiscoveryAssistant.com. One of our team will be in touch. If you’re interested in doing a free trial of our case law and resource database, please jump to eDiscoveryAssistant.com and sign up to get started or drop us a line at support@eDiscoveryAssistant.com.
Have a great week. Stay safe and healthy. Hug your loved ones close if you can, and I’ll see you next week. Thank you.