Salesforce logged four separate incidents on 25 September 2026, running from 02:30 UTC through to 20:50 UTC, touching Discovery search, Marketing Cloud Engagement logins, Commerce Cloud traffic dashboards and the Intelligence service. All four were faults in Salesforce's own infrastructure, with the longest running 11h 58m.
Started
18:51 UTC
Duration
Lasted 1h 59m
Source
IsDown
Next time
Salesforce is degraded. Is your product working?
An incident on their side does not always mean an outage on yours, and the only way to know is to already be watching your own endpoints. Set that up in under a minute and you will be notified the next time your own site is affected. And there will be a next time.
50 monitors for your own websites. No card.
Affected components: POD217, POD240, POD248, POD244, POD253, POD256, POD258, POD267, POD270, POD284, POD297, POD304, POD306, POD311, DB401000, DB401001, DB401002, DB401003, DB401004, DB401005, DB401006, DB401008, DB401009, DB401011, DB401012, DB401013, DB401014, DB401016, DB401017, AUS2S, JPN2S, JPN4S, CAN2S, AUS4S, AUS22S, AUS24S, AUS26S, AUS6S, USA1, USA2S, USA3S, DEU2S, DEU4S, BRA2S, SGP2S, FRA2S, FRA4S, CAN6S, CAN8S, JPN10S, JPN12S, AUS36S, IND22S, DEU6S, JPN18S, JPN24S, AUS16S, AUS18S, AUS20S, FRA6S, FRA12S, FRA24S, FRA32, SGP10, BRA6S, BRA8S, DEU10S, DEU12S, DEU14S, DEU16S, CAN26, CAN28, CAN20S, FRA34, DEU36, DEU38, DEU20S, USA32S, FRA36, GBR2S, SWE2S, NA223, NA225, NA224, CAN34, AUS58, AUS60, AUS62, AUS64, DEU52, ARE6, AUS66, AUS68, AUS70, AUS72, AUS76, AUS82, AUS84, AUS86, AUS88, AUS90, AUS92, BRA14, BRA16, BRA18, CAN36, CAN38, CAN40, CAN42, CAN44, CAN46, CAN48, CAN50, CAN52, CAN54, CHE2S, CHE4S, CHE6, DEU54, DEU56, DEU58, GBR34, GBR36, GBR38, GBR40S, GBR42S, GBR44S, GBR46, GBR48, GBR50, GBR52, GBR54, GBR58, GBR62S, GBR64, GBR68, GBR70, GBR74, IDN6, ITA2S, ITA4S, ITA6, JPN132, JPN134, JPN140, JPN142, NA232, NA233, NA235, NA236, NA237, NA238, SGP14, SWE10, SWE12, SWE14, SWE16, SWE18, SWE20, SWE22, SWE24, SWE26, SWE28S, SWE30S, SWE32, SWE34, SWE36, SWE38, SWE40, SWE42, SWE46, SWE6S, SWE8, USA224S, USA226, USA228, USA240S, USA242S, USA244S, USA250S, USA252S, USA258S, USA260S, USA286, USA294, USA302, USA304, USA312, USA314, USA316, USA320, USA322, USA324, USA326, USA328, USA330, USA332, USA334, USA336, USA338, USA342, USA344, USA346, USA348, USA350, USA352, USA354, USA356, USA360, USA366S, USA372, USA374, USA376, USA384, USA412, USA416, USA418, USA422, USA424, USA428, USA430, USA438, USA440, USA442, USA444, USA452, USA456, USA462, USA466, USA468, USA470, USA474, USA476, USA478, USA480, BRA24, BRA28, BRA32, BRA34, BRA36, BRA38, BRA44, BRA48, BRA50, CAN56, CHE12, CHE16, CHE18, CHE26S, CHE30, DEU106, DEU108, DEU110, DEU112, DEU118, DEU120, DEU124, DEU126, DEU128, DEU66, DEU68, DEU70, DEU72, DEU74, DEU76, DEU78, DEU84, DEU86, DEU92, DEU94, DEU96, EU53, FRA100, FRA42, FRA44, FRA46, FRA48, FRA52, FRA54, FRA56, FRA66, FRA76, FRA78, FRA86, FRA88, FRA90, FRA96S, FRA98S, GBR100, GBR102, GBR108S, GBR110S, GBR112, GBR116, GBR122, GBR76, GBR80, GBR82, GBR90S, GBR94, GBR96, GBR98, IND134, IND136, IND138, ISR6, ITA10, ITA12, ITA14, ITA16, ITA18, ITA30, ITA36, ITA38, ITA42, ITA8, JPN146, KOR6, NA239, NA240, NA241, NA242, NA243, NA244, NA246, NA247, NA248, NA249, NA251, NA252, NA253, SWE106, SWE114, SWE118, SWE122, SWE48, SWE50, SWE52, SWE54, SWE56, SWE58, SWE60S, SWE62, SWE64, SWE66, SWE72, SWE74, SWE76, SWE78, SWE80, SWE82, SWE84, SWE88S, SWE90S, SWE92S, SWE94S, SWE96S, SWE98, USA594, USA622, USA656S, USA658S, USA702, USA708, USA710S, USA712S, USA714, USA716, USA718, USA722, USA724, USA726, USA728, USA730, USA732, USA736, USA738, USA740, USA742, USA744, USA746, USA748, USA750S, USA754S, USA756S, USA758S, USA760S, USA762S, USA764S, USA766S, USA768S, USA770S, USA772S, USA774S, USA776, USA778, USA780, USA786S, USA788, USA792S, USA810, USA838, USA840, USA844, USA846, USA856, USA864, USA866, USA868, USA870S, USA872S, USA874S, USA876, USA9002, USA9014, USA916S
What happened?
The day opened at 02:30 UTC with a Feature Disruption traced to a certificate expiry that broke requests to the Discovery service, causing search and index rebuild failures for a subset of B2B Commerce customers. Salesforce's updates show the impact radius growing twice during that incident as more instances were found affected, and the fix involved generating and rotating certificates region by region, validating each before moving to the next. That incident ran 3h 58m. At 08:55 UTC a second incident hit Marketing Cloud Engagement, blocking logins to Stack 401. Salesforce identified a high volume of login requests from a small number of integrations, applied rate limits, and confirmed there was no underlying database fault. That one lasted 1h 48m. The longest incident of the day started at 11:29 UTC and ran 11h 58m: Commerce Cloud customers saw zero values on their site visit traffic dashboards, even though Salesforce confirmed the underlying data was being recorded correctly. The cause was a blocked processing pipeline. A configuration change unblocked it by 13:44 UTC, but Salesforce said the backlog would take several hours to fully populate, and as of the 19:55 UTC update it was still catching up. A fourth incident began at 18:51 UTC, affecting the Intelligence service and disrupting logins. Salesforce's own note calls it a recurrence of an earlier US2 incident, with the US2 Redis cache again hitting its memory limit and rejecting writes. That ran 1h 59m. Across all four, 12 of Salesforce's 416 listed components were marked affected, and no user reports were filed into this record during the window.
Learning
None of these four incidents were the kind of thing an external uptime check against your own site would catch, because the faults sat inside Salesforce's infrastructure: an expired internal certificate, a login rate-limiting problem triggered by one integration, a blocked data pipeline, and a Redis cache hitting its memory limit. Your own endpoints could have stayed green the whole time while your customers saw empty traffic dashboards or failed logins, because the failure was in Salesforce processing or serving your data correctly, not in whether your site responded. Salesforce's own pattern here is worth noting too: the Redis cache problem that hit the Intelligence service at 18:51 UTC was flagged by Salesforce itself as a recurrence of an earlier issue, which tells you single fixes do not always hold. Across the last 90 days Salesforce has logged 127 incidents with a median duration of 2 hours 31 minutes, so a day with four incidents and one running close to 12 hours is not unusual for this provider. Your monitoring covers your half, UptimeRobot now watches the provider's half too, so next time you know which side broke without guessing.
