Advisory: Reflexion - Delayed delivery of outbound and inbound email

[Update] 7/31/2019 - 7:15AM EDT

As of 6:30 AM EDT, Reflexion services continue to process the backlog of email that was delayed as a result of the outage yesterday.

• No email was lost, and we expect all queued email to be delivered over the next few hours.

• Our infrastructure engineering teams continue to monitor and are actively working to resolve the issue and restore regular mail flow as quickly as possible.

• Reflexion Outbound mail is delivering without issue.

• Again, we sincerely apologize for the inconvenience this has caused and will provide an update once the queues return to normal or there is additional information to share.

[Update] 7/31/2019 - 7:15AM EDT

As of 6:30 AM EDT, Reflexion services continue to process the backlog of email that was delayed as a result of the outage yesterday.

• No email was lost, and we expect all queued email to be delivered over the next few hours.

• Our infrastructure engineering teams continue to monitor and are actively working to resolve the issue and restore regular mail flow as quickly as possible.

• Reflexion Outbound mail is delivering without issue.

• Again, we sincerely apologize for the inconvenience this has caused and will provide an update once the queues return to normal or there is additional information to share.

Up at 5a to deal with this issue again here on west coast.  Already getting emails from my customers at 4:30a replying to emails I sent late last night.  Sent email tests to two customers at 5:18a and as of 5:54a they are not delivered nor are they in the queue waiting to be delivered.  I've moved 4 customers off of Reflexion last night bypassing via MX records.  If the delay is still 45-60+ minutes for inbound emails I will need to spend all morning updating MX records.  

Also, there is still no communication regarding the actual issue.  How is that possible?  Who is in charge of Reflexion and why is there such poor communication even as of this morning?

If your update was at 7:15a eastern and you expect the queues to be caught up within a few hours then we are approaching or have surpassed this timing.  What's the status of the queues?  How long is it taking for emails to be delivered?  How long are the delays?  This type of actionable information allows me to make decisions.  All of your updates so far are generic and making our job more difficult.  I can't expect my customers to wait an hour or who knows how long for emails which are normally delivered within seconds.  In their eyes they rightfully will believe the system is still not working.

Now 6:10a PST and multiple test emails to customers not delivered and not showing in queues.  Approaching one hour.  They are receiving some emails albeit one here and there and who knows when those were actually sent to them.

Engineering is sending more generic updates this morning.  Where is management with any communication?  What was the reason for Sophos purchasing Reflexion only to abandon it?  I was a huge fan of Sophos and Reflexion but now it looks like Reflexion has major infra issues and Sophos has shown that during a crisis they have no clue how to communicate.

For anyone looking for actionable information.  I sent test emails from O365 (with Reflexion filtering turned OFF in both directions) to Reflexion customers at 5:18a.  They are not even showing up in the queues as of 7a Pacific.  That's 1.75 hour delay at a minimum.  I'm in process of updating MX records to skip Reflexion as this is not an acceptable delay.  I've left the outbound connector alone for now as they appear to be timely.

This matches what we're seeing too.  Messages that happen to reach Reflexion are being delivered and outbound seems to be working.  But Reflexion isn't accepting most of the incoming mail that's queued up for it.  Trying to find a common denominator.

What is odd, some messages are coming through immediately.

7/31/19 - 2:30PM EDT - Update on Issues with Reflexion Email Security Service:

• As of 2:30 pm EDT, all inbound and outbound mail delivery through Reflexion is returning to normal, although some end users may periodically notice delays due to the high volume of emails we are currently processing. We expect that any delays will continue to improve as the queues are processed. We will continue to provide updates as the operation progresses back to a fully healthy state with no delays.
• Please know that all email that was queued in the last 24 hours has been safely processed and will be sent and received as the flow resumes - no emails were lost by the Reflexion operation as a result of this issue.
• We expect improvements throughout the day. The operations and development teams are diligently monitoring the service to ensure all Reflexion customer email is fully resumed. Note: This issue affected customers of the Reflexion service and did not impact any of our services in Sophos Central.
• While we are still finalizing the root cause of the problem, we have determined that during the initial incident the IO limits of our storage infrastructure was exceeded, due to a routine maintenance process that was executed incorrectly. We are continuing the investigation and will provide a thorough root cause analysis (RCA) once complete.
• We sincerely apologize for the inconvenience this has caused and will continue to provide updates during the day. Please continue to watch this KBA page.

https://community.sophos.com/kb/en-us/134470

Any info on queue sizes?  I'm still seeing delays on the couple of customers I have still using Reflexion.  Everyone else is bypassing it for now.  I'd like to set everyone back but I don't want that to do that until I know email is back to normal.

7/31/19 - [Update] 9:00 pm EDT:

• All Reflexion services have been completely restored, and all queued emails have been delivered. The infrastructure and engineering teams continue to diligently monitor the service. 
• While we are still finalizing the root cause of the problem, we have determined that during the initial incident the IO limits of our storage infrastructure were exceeded. 
• We sincerely apologize for the inconvenience this has caused and are continuing the investigation. We will provide a thorough root cause analysis (RCA) once it’s complete. 

8/1/19 - 9:00 am EDT - Update:

• All Reflexion services have been completely restored, and all queued emails have been delivered. The infrastructure and engineering teams continue to diligently monitor the service.
• While we are still finalizing the root cause of the problem, we have determined that during the initial incident the IO limits of our storage infrastructure were exceeded. 
• We sincerely apologize for the inconvenience this has caused and we will provide a thorough root cause analysis (RCA) once it’s completed. 
• We will continue to provide updates as there is additional information around this incident to report.