Linode Status · History · Incident #6089

RESOLVED

Upstream Loss - Some Paths (in-maa/in-bom-2 to US Region)

Minor · Started Jul 31, 2026 · 5:49 PM

  • Duration

    4d 7h 45m

  • Severity

    Minor

  • Detection lead

  • User reports

Summary

Upstream Loss - Some Paths (in-maa/in-bom-2 to US Region)

On July 31, 2026, at approximately 10:00 UTC, Akamai observed intermittent network losses affecting compute users accessing US locations from our India sites \(MAA and BOM\). Customers’ services in North America, particularly the Miami data center region, experienced increased latency, intermittent connectivity issues, slower data transfers, and difficulty reaching certain applications or services. Performance was unstable, with periods of normal operation followed by disruptions. To address the issue, Akamai applied a deny-all policy to the impacted upstream provider transit link, redirecting traffic around the impacted routes. Despite this mitigation, ongoing IPv6 losses occurred due to congestion between two alternative upstream providers, impacting some users. One provider acknowledged a bottleneck in the Asia region, and the alternate provider worked to reroute traffic away from affected links.  The initial impacted service provider confirmed that two fiber cuts in Mexico caused congestion on the impacted routes. One of these fiber cuts was resolved at 23:43 UTC on July 31, 2026, and no further issues were observed following this mitigation.  This summary provides an overview of our current understanding of the incident given the information available. Our investigation is ongoing and any information herein is subject to change.


  • Started

    Jul 31, 2026 · 5:49 PM

  • Resolved

    Aug 5, 2026 · 1:34 AM

  • Duration

    4d 7h 45m

  • Severity

    None

Event timeline

How this incident unfolded

  • Investigating

    Jul 31 · 5:49 PM Linode

    Our team is investigating packet loss on some paths from the in-maa & in-bom-2 data centers into the US region. During this time, users may experience connection timeouts and errors with services traversing between India and United States regions. We will share additional updates as we have more information.

  • Identified

    Jul 31 · 6:25 PM Linode

    Our team has identified the cause of packet loss to our US data centers. We are working with our upstream provider to resolve this issue and we will provide an update as soon as the solution is in place.

  • Identified

    Aug 4 · 6:36 PM Linode

    We are continuing to work with our upstream provider to resolve the packet loss affecting some paths between the in-maa and in-bom-2 data centers and the US region. We will share further updates as progress continues.

  • Resolved

    Aug 5 · 1:34 AM Linode

    At this time the upstream provider has been able to correct the issue causing packet loss on some routes from the in-maa & in-bom-2 data centers into the US region and will now consider this incident resolved. If you continue to experience problems, please <a href="https://cloud.linode.com/support/tickets">open a Support ticket</a> for assistance.

  • Postmortem

    Aug 5 · 5:49 PM Linode

    On July 31, 2026, at approximately 10:00 UTC, Akamai observed intermittent network losses affecting compute users accessing US locations from our India sites \(MAA and BOM\). Customers’ services in North America, particularly the Miami data center region, experienced increased latency, intermittent connectivity issues, slower data transfers, and difficulty reaching certain applications or services. Performance was unstable, with periods of normal operation followed by disruptions. To address the issue, Akamai applied a deny-all policy to the impacted upstream provider transit link, redirecting traffic around the impacted routes. Despite this mitigation, ongoing IPv6 losses occurred due to congestion between two alternative upstream providers, impacting some users. One provider acknowledged a bottleneck in the Asia region, and the alternate provider worked to reroute traffic away from affected links.  The initial impacted service provider confirmed that two fiber cuts in Mexico caused congestion on the impacted routes. One of these fiber cuts was resolved at 23:43 UTC on July 31, 2026, and no further issues were observed following this mitigation.  This summary provides an overview of our current understanding of the incident given the information available. Our investigation is ongoing and any information herein is subject to change.

Get alerted before the next Linode outage.

Pulsetic catches degradations minutes before vendors acknowledge them.

Start monitoring free