## The symptom
An Astra DB Serverless database is stuck in HIBERNATED state. `astra db resume`
(Astra CLI v1.1.1) fails with:
```
An unexpected error occurred during the execution of the command (IOException):
HTTP/1.1 header parser received no bytes, caused by java.io.EOFException:
EOF reached while reading
```
The same database also fails to resume through the Astra Portal, and a direct
Data API request returns `curl: (52) Empty reply from server`.
## Why
This was backend degradation on Astra's side, not a CLI bug. The control plane
was closing the connection before returning HTTP headers when resuming
hibernated databases, so the CLI's HTTP parser got zero bytes. The maintainer
confirmed the degradation.
## How to tell it is backend, not you
- The token is valid: the CLI can fetch org info and database status fine.
- The portal fails too, not just the CLI.
- A direct Data API request returns an empty reply.
Do not burn time debugging your token. Check the Astra status page and retry
later. Once the service was restored, the reporter re-ran `astra db resume` and
the database transitioned HIBERNATED -> RESUMING -> ACTIVE, confirmed in the
portal.
## If resume keeps failing while status calls succeed
Open a support case. The backend may be stuck on your database specifically.
## Separate issue worth knowing
`astra setup` failing with `NoSuchElementException: No value present` on
v1.1.1 is a distinct CLI bug, fixed in v1.1.3. Pass the token directly with
`--token @file` as a workaround on the older version.