Inside CIOs’ response to the CrowdStrike outage — and the lessons they learned

Mike Mainiero was awakened at around 2 a.m. on July 19 by someone on his incident response team. Catholic Health’s IT systems and those of its partners were crashing, with one of its radiology vendors being among the first to experience technical difficulties.

With nearly 16,000 employees, the Long Island-based healthcare system supports six acute care hospitals, three nursing homes, a home health service, a hospice, and a network of physician practices. It has a triage team that is set up to assemble IT staff and establish a command center for cybersecurity and other incidents, says Mainiero, Catholic Health’s SVP and CDIO.

“We needed to muster teams immediately to first try and understand what was happening and then triage, communicate, and mitigate,” he says, adding that officials quickly realized this was a priority-one incident. “Obviously, when anything is happening in a hospital, people can die.”

Like Mainiero, many CIOs got a rude awakening on July 19 with the news that their organization had experienced an outage due to a faulty software update from cybersecurity company CrowdStrike. Millions of computers globally running Microsoft Windows crashed and displayed the “blue screen of death” error message.

Mainiero then began messaging the C-suite to apprise them of the situation and let them know the command center had been set up.

Catholic Health received notification from CrowdStrike with information about the remediation, which had to be applied manually to hundreds of servers, Mainiero says. That was followed by remediation of desktops and workstations in patient care settings.  

The command center provided touchpoint updates with operational leads at all the healthcare system’s facilities. “We were able to mitigate this so we didn’t have to cancel any surgeries or appointments,’’ and their facilities were still able to provide care to patients, with staff taking notes on paper, he says. All critical issues were mitigated by 5 p.m. that day.

If the outage had occurred during the day, the impact would have been greater, Mainiero notes. “In a hospital setting, you don’t want to divert patients to another setting, and we were able to avoid all diversions,’’ he says. “So it definitely impacted the end users but not the patients. Because of our quick reaction, we were able to mitigate any harm.”

All-hands response

While not all life-threatening situations, this same scenario played out at organizations across the globe. For many CIOs, being prepared, having disaster recovery and business continuity plans in place, and constant communication with stakeholders lessened the impact of the outage.

“We were able to get all servers up and running within three hours and … all laptops up and running by that Friday afternoon,’’ says John Roman, CIO at The Bonadio Group, a national CPA firm, estimating that about 300 out of 1,100 total devices were impacted.  “The reason we were able to do that was we implemented our incident response plan. Most incident response plans are created in the event there’s some type of malware incident. We genericized ours to take into consideration any type of incident — including a global pandemic.”

Once the incident response plan was deployed, the second step was calling on everyone in IT to implement the script CrowdStrike created to fix the problem, says Roman, who was also in constant communication with the firm’s leaders and all employees through their firmwide texting service, intranet portal, and email.

Like Mainiero, Roman says the timing of the outage made a difference. “If there’s any good news, it’s that we’re an accounting firm and we’re busy all year round, however, probably our busiest time of year is tax season,’’ he says. “Had this happened in March, there would have been a significant business impact because we wouldn’t have been able to perform tax services, but because it happened in the dead of summer and we were able to remediate it as fast as we did, the business impact was minimal.”

The impact was more significant for Black Wallet, a stablecoin 2.0 ecosystem management company. CIO Remi Alli characterizes it as “a challenging experience for our organization,’’ and says the outage lasted for several hours, “during which we were unable to access critical security functionalities, impacting both our servers and laptops.”

Without access to CrowdStrike’s services, Alli says, “we were unable to effectively monitor and respond to potential threats, which raised concerns about our overall security posture. Overall, it was a tense and challenging time as we had to work around the limitations imposed by the outage while maintaining the integrity of our systems.”

Staying calm is key

The first thing Alli did was gather the incident response team to assess the situation and establish the company’s immediate response plan. “We had to ensure that we could maintain business continuity while we addressed the implications of the outage,’’ Alli says.

Communication was vital and Alli kept leadership and stakeholders informed about the situation and the steps IT was taking with regular updates. “It’s easy to panic in these situations, but we focused on being transparent and calm, which helped to keep the team grounded,’’ Alli says.

Additionally, “The lack of access to critical security insights put us at risk temporarily, but more importantly, it highlighted vulnerabilities in our overall security posture. We had to quickly shift some of our security protocols and rely on other measures, which was a reminder of the importance of having a robust backup plan and redundancies in place,’’ Alli says.

Mainiero agrees, saying that in this type of situation, “you have to take on a persona — if you’re panicked, your teams are going to panic.” He says that training has taught him never to raise his voice. “But I’m going to have a command-and-control tone … I might be firm [and] unapologetic, but always kind. You want to inspire people because they’ll essentially go above and beyond for you if you don’t bark orders.”

Lessons learned and other takeaways

Accounts payable software provider AvidXchange saw a portion of its customer-facing product portfolio impacted by the outage, but CIO Angelic Gibson says IT was able to restore services completely in less than 24 hours. She credits this with being “overly prepared” and having scenario plans in place in the event of an outage.

“We acted in accordance with our business interruption plan,’’ Gibson says. “Taking steps to have proactive preparation and having the right communications channel in place allowed us to communicate quickly and effectively to all necessary parties while we worked internally to get our systems back up and running.”

Gibson says CIOs must communicate openly and often, as well as show confidence in the playbook and provide reassurance, ensure they have redundancy, and be prepared to mobilize resources so IT can act quickly.  

Catholic Health’s Mainiero emphasizes that having a pre-built communication plan is paramount, with lists of people and a platform that lets you create canned subject lines and texts. Similarly, it’s important to have a system that lets you “automate mustering your teams to assemble.”

He also recommends having vendor contacts on call. Even though Mainiero dealt directly with CrowdStrike, he says it was important to notify everyone affected by the outage, given that Catholic Health works with hundreds of vendors.

Like the others, The Bonadio Group’s Roman says an organization’s incident response plan needs to take into consideration any major incident — not just a ransomware or malware attack. Further, it’s not enough to just have a plan in place; you have to practice it.

“We do tabletop exercises annually and based on those we’ll update our incident response plan,’’ Roman says.

Depending on the size of your IT department, it should be all hands on deck if there’s a company-wide outage.

“Then communicate, communicate, and communicate,’’ Roman says. “People want to know what’s going on throughout an outage. I travel a lot and one thing that’s probably the most aggravating is when a flight gets canceled and you don’t know why. But when you do know why, you don’t feel great … but you might feel a little better.”

Black Wallet’s Alli learned several key lessons from the outage. The first was to “prepare for the unexpected. No matter how reliable a vendor seems, always have contingency plans,’’ Allie says. “Testing our incident response protocols for a range of scenarios, including vendor outages, is now a priority.”

Another is to strengthen internal communication not just within IT, but across the entire organization. This can help mitigate panic and ensure everyone is on the same page, Alli says.

The outage also prompted “a deep dive into our third-party dependencies, reviewing not just CrowdStrike but all critical vendors. Understanding our reliance on them has led to more proactive management of those relationships and better risk assessments.”

It also reinforced the need for Black Wallet to cultivate strong cyber hygiene practices within the organization, Alli says.

Post-recovery, Black Wallet “conducted a thorough post-mortem to analyze not just what went wrong but how we responded,’’ Alli says. “This reflective practice will help us improve our processes moving forward.”

Overall, while the CrowdStrike outage was a tough experience, it reminded the organization of “the importance of resilience and adaptability in our approach to cybersecurity,’’ Alli says. “The recovery process has involved not just technical fixes but also a strengthening of our organizational culture around security and risk management.”

Remaining loyal

Both Roman and Mainiero are emphatic that the outage will not impact their relationship with CrowdStrike.

Despite the criticism the company has received, “CrowdStrike is an amazing company, and we look at [vendors] for their culture of engineering,’’ Mainiero says. “They made a mistake.” Healthcare systems have to be prepared for anything to go down, he adds.

Noting that CrowdStrike has been accused by some of having a monopoly in the industry, Mainiero says that this incident was a quality assurance issue. With systems and integrations becoming more sophisticated and complex, he says that the industry should come together to discuss how to handle quality assurance.

Roman echoed that, saying, “We’re all human beings. We all make mistakes. If you’re going to have the attitude of abandoning CrowdStrike … you’d have to consider abandoning any cloud-based provider. Up to this point, CrowdStrike has worked flawlessly and has prevented virus and malware outbreaks, their support has been outstanding, and we’re still committed to our relationship with them.”