Xendit recorded a dense series of payment disruptions across Indonesia and the Philippines on September 30, with LinkAja failing repeatedly while GCash, DANA, QRPH, ECPay and Philippine payout routes experienced separate periods of degraded performance.
The incidents did not amount to a platform-wide Xendit outage. The company continued to mark its core API, cards and most major services as operational on its official status page. Instead, the failures appeared across individual payment partners and routes, with Xendit repeatedly attributing them to disruptions on the partner side.
That distinction matters because Xendit sits between merchants and a fragmented network of e-wallets, banks, QR systems and cash-payment providers. A merchant may integrate with one payments platform, but the transaction still depends on multiple underlying financial systems working correctly.
LinkAja Failed Repeatedly Across E-Wallet and QR Payments
LinkAja generated the most persistent pattern of incidents during the day.
Xendit logged an Indonesian LinkAja e-wallet disruption at 09:35 WIB, followed later by a succession of failures affecting both e-wallet and QR-code payments. New incidents were recorded at approximately 12:55, 13:16, 13:32 and 14:03 WIB.
The failures returned again late in the afternoon. Xendit identified another LinkAja e-wallet disruption at 17:10 WIB, followed by a QR-code incident five minutes later. Further e-wallet trouble appeared at 17:38, with another QR-code disruption recorded at 18:06.
In the incidents, Xendit said LinkAja’s successful payment rate had fallen below its normal threshold. Merchants could still initiate transactions, but payment-completion callbacks could be delayed. Xendit recommended directing customers toward alternative e-wallet providers while the problems persisted.
For affected transactions, the company also warned that reconciliation would be required after service recovered.
That is an important operational detail. A delayed callback does not necessarily mean the customer’s payment failed. It can mean the payment succeeded at one layer while the merchant has not yet received confirmation, creating the risk that customers retry transactions or merchants temporarily treat completed payments as unpaid.
GCash and DANA Saw Sharp Callback Delays
LinkAja was not the only e-wallet experiencing problems.
DANA, another major Indonesian wallet, suffered elevated callback latency early on September 30. Xendit measured delays reaching 124.643 seconds during the previous 10-minute monitoring window before marking the incident resolved at 08:29 WIB.
GCash in the Philippines then experienced several separate callback-latency incidents. Xendit recorded peaks of 21.753 seconds, 25.027 seconds and 24.561 seconds across three disruptions during the morning.
In each case, transaction initiation remained available, but confirmation could arrive later than expected.
For consumers, a delay measured in tens of seconds may sound minor. For high-volume digital merchants, however, callback reliability is part of the transaction itself. Checkout pages, order systems, fraud engines and inventory software often depend on rapid confirmation before advancing a purchase.
The problem becomes more serious when many financial services are chained together. Dave Finances recently examined how payment delays can become reconciliation problems when different institutions show different stages of the same transaction as complete.
QRPH Returned 503 Errors While Philippine Payouts Slowed
Separate problems affected other Philippine payment rails.
Xendit reported a QRPH disruption at 05:23 WIB in which some QR payment charge requests encountered intermittent timeouts. Affected merchants received HTTP 503 responses with the error code CHANNEL_UNAVAILABLE.
Xendit said more than 0.118% of requests during the previous 15 minutes had been affected. Unlike the e-wallet callback incidents, the company advised merchants to retry transactions receiving the 503 error and said post-incident reconciliation would not be necessary.
Earlier in the day, Philippine payouts were also delayed between 00:40 and 01:50 WIB.
The disruption affected transfers to UnionBank of the Philippines above PHP 50,001, while transfers between PHP 1 and PHP 50,000 to other banks and e-wallets were also delayed. Xendit again attributed the issue to a partner system.
The episode illustrates why companies building global payment infrastructure around local rails spend so much effort adding banking partners and alternative routes. The API that a merchant sees may be unified, but the networks underneath it remain highly fragmented.
ECPay Loan Payments Failed Several Times
Xendit also logged recurring disruptions affecting ECPAY_LOAN, an over-the-counter payment channel in the Philippines.
During the incidents, customers could still create or update payment codes, but newly created codes could not be paid. Successful over-the-counter payments could also take longer than normal to process.
One disruption that began on September 29 remained unresolved until 07:09 WIB on September 30. Another appeared at 09:07 and was resolved at 09:45. A further disruption began at 10:39 and lasted until 14:34.
By 15:07 WIB, Xendit had opened another ECPAY_LOAN incident with the same basic symptoms.
The recurrence makes the ECPay problem more notable than a single short interruption. It suggests that restoring service once did not necessarily eliminate whatever condition was causing the partner channel to fail.
There Is No Evidence Yet of One Shared Failure
The obvious question is whether so many disruptions occurring on the same day point to one common infrastructure problem.
For now, the public evidence does not support that conclusion.
LinkAja and DANA operate in Indonesia, while GCash, ECPay and QRPH belong to the Philippine payments ecosystem. The incidents also affected different products: e-wallet callbacks, QR charge creation, over-the-counter payment codes and bank or wallet payouts.
Xendit classified the problems separately and repeatedly described them as disruptions on its partners’ systems. Its broader platform remained largely operational.
That makes several independent partner failures a plausible explanation. Another possibility is that some routes share an intermediary processor, switch, cloud provider or other infrastructure dependency that has not been publicly identified.
But without a postmortem, partner disclosure or matching technical evidence, connecting the incidents would be speculation.
The Bigger Risk Is Payment Infrastructure Concentration
The more useful takeaway is not that one invisible provider necessarily broke everything. It is that modern payments can appear consolidated at the merchant interface while remaining fragmented underneath.
A business using a single gateway may still depend on half a dozen wallets, banks, processors, QR networks and settlement providers. That architecture creates convenience during normal operations, but it can also hide where the real operational dependencies sit.
This is becoming more important as payment companies increasingly compete on orchestration. Visa, Mastercard and Stripe, for example, are investing heavily in the ability to route different forms of money through increasingly complex payment infrastructure. The value of that orchestration depends not only on adding more rails but on being able to detect failures and redirect transactions when one rail stops working.
The same challenge applies to newer Asian payment networks. Companies such as dtcpay are expanding regulated payment infrastructure across Southeast Asia, where connecting more local rails creates wider reach but also increases the number of counterparties that can affect service quality.
What Merchants Should Watch After September 30
The most important next signal will be whether the recurring LinkAja and ECPay failures continue into October or disappear after partner-side fixes.
If the incidents stop, September 30 may ultimately look like an unusually noisy day across several unrelated payment networks.
If they continue, the pattern becomes more significant.
Repeated restoration followed by another outage can indicate an unresolved upstream problem rather than a clean one-time failure. Merchants would then need to pay closer attention to alternative routing, callback handling, retry logic and automated reconciliation.
The customer experience is also critical. A payment that fails clearly is inconvenient. A payment that succeeds but remains unconfirmed is potentially worse because neither the customer nor the merchant immediately knows whether another attempt is safe.
That is why the callback incidents matter more than their short durations suggest.
For Xendit, the September 30 cluster is ultimately a test of what payment orchestration is supposed to provide. Connecting merchants to dozens of local financial networks creates value because businesses do not have to manage every relationship themselves. But when several of those networks degrade at once, the aggregator also becomes the place where merchants expect visibility, failover and reliable reconciliation.
The unanswered question is whether September 30 was simply several partners having a bad day — or an early sign of a shared infrastructure weakness that has not yet been publicly identified.
Johan Shamshad is a financial markets writer at Dave Finances covering cryptocurrencies, trading platforms, brokers, fintech, financial regulation, and developments across global markets. He previously worked at Gulf News, adding newsroom experience to his coverage of fast-moving financial and digital-asset markets.
His work focuses on identifying market-moving events, company developments, regulatory changes, product launches, and shifts in trading and financial infrastructure.
Johan contributes news and analysis designed to help readers understand not only what happened, but why a development matters and how it may affect the wider financial landscape.