Bug Report: Outgoing messages stuck in PENDING, never delivered (Evolution API 2.4.0, Baileys)**
Evolution API version
2.4.0 (homolog tag), license activated successfully (community tier)
Database
PostgreSQL 15
Connection type
Baileys (WhatsApp Web)
What I did
Sent text messages via /message/sendText/{instance}, both from my own backend and directly from the Evolution Manager UI, to a real, active WhatsApp number that I personally control and actively monitor.
What happened
- API call returns 200 OK with a valid messageId, status PENDING
- Message never progresses past PENDING in Message.status
- Message is never received on the destination phone (confirmed visually on the device, multiple times, over several hours)
- This happens regardless of: client (own backend vs Manager UI directly), message content, or time of day
Database evidence
sqlSELECT status, COUNT(*) FROM "Message" WHERE "key"->>'fromMe'='true' GROUP BY status;Result: 74 PENDING, 1 READ, 1 SERVER_ACK — out of 76 total outbound messages since instance creation (June 18). This has been consistent since day one, not something that started recently.
Instance state
connectionState reports state: "open" consistently and stably — instance shows as connected.
Logs (evolution-api container)
Repeated, on every message send attempt:Closing session: SessionEntry { ... }On first pairing (once): a stream:error with code: "515", followed by an automatic full session restart (ChannelStartupService re-init).
Steps already taken (none resolved it)
- Full logout + re-scan QR from scratch — same behavior after reconnect
- Updated from 2.3.7 (latest) to 2.4.0 (homolog) + activated community license — same behavior
- Confirmed correct webhook config (was a separate, now-fixed issue: webhookByEvents: true was causing 404s on event-specific routes — fixed, unrelated to this delivery issue)
- Sent directly via Manager UI (bypassing my own backend entirely) — same PENDING-forever result
- Verified the receiving number is correct and active (confirmed visually on the device)
Possible related context
The conversation in question appears to use a WhatsApp @lid JID (remoteJid ending in @lid with remoteJidAlt as the real phone number) for inbound messages — possibly related to WhatsApp's newer privacy/multi-device addressing mode. I'm aware of the documented @lid onWhatsApp exists:false issue, but my outbound sends to the plain @s.whatsapp.net JID (the phone number) are equally stuck in PENDING, so I'm not certain this is the same root cause.
Question
Is PENDING-forever with no SERVER_ACK/DELIVERY_ACK a known symptom of something specific (rate limiting, account flagging, a particular WhatsApp Web version mismatch)? Is there a way to get more diagnostic detail on why a specific message never leaves PENDING (e.g. a stream-level error per message, not just connection-level)?
Happy to provide instance_id, full logs, or any other diagnostic info needed.
Bug Report: Outgoing messages stuck in PENDING, never delivered (Evolution API 2.4.0, Baileys)**
Evolution API version
2.4.0 (homolog tag), license activated successfully (community tier)
Database
PostgreSQL 15
Connection type
Baileys (WhatsApp Web)
What I did
Sent text messages via /message/sendText/{instance}, both from my own backend and directly from the Evolution Manager UI, to a real, active WhatsApp number that I personally control and actively monitor.
What happened
Database evidence
sqlSELECT status, COUNT(*) FROM "Message" WHERE "key"->>'fromMe'='true' GROUP BY status;Result: 74 PENDING, 1 READ, 1 SERVER_ACK — out of 76 total outbound messages since instance creation (June 18). This has been consistent since day one, not something that started recently.
Instance state
connectionState reports state: "open" consistently and stably — instance shows as connected.
Logs (evolution-api container)
Repeated, on every message send attempt:Closing session: SessionEntry { ... }On first pairing (once): a stream:error with code: "515", followed by an automatic full session restart (ChannelStartupService re-init).
Steps already taken (none resolved it)
Possible related context
The conversation in question appears to use a WhatsApp @lid JID (remoteJid ending in @lid with remoteJidAlt as the real phone number) for inbound messages — possibly related to WhatsApp's newer privacy/multi-device addressing mode. I'm aware of the documented @lid onWhatsApp exists:false issue, but my outbound sends to the plain @s.whatsapp.net JID (the phone number) are equally stuck in PENDING, so I'm not certain this is the same root cause.
Question
Is PENDING-forever with no SERVER_ACK/DELIVERY_ACK a known symptom of something specific (rate limiting, account flagging, a particular WhatsApp Web version mismatch)? Is there a way to get more diagnostic detail on why a specific message never leaves PENDING (e.g. a stream-level error per message, not just connection-level)?
Happy to provide instance_id, full logs, or any other diagnostic info needed.