Redis
WRONGTYPE Operation against a key holding the wrong kind of value
WRONGTYPE Operation against a key holding the wrong kind of value: what causes it, how to check a key's real type, and how to fix the command.
What it means
Redis returns WRONGTYPE Operation against a key holding the wrong kind of value when a command is run against a key whose stored data type doesn’t match what that command expects — for example, calling LPUSH on a key that already holds a string, or HGET on a key that’s actually a set. Every key in Redis has exactly one type at a time, and Redis enforces that strictly rather than silently coercing the data. (SET is the one common exception: it overwrites a key of any type, discarding the old value.)
Common causes
- The key already exists with a different type, often because an earlier
SETcreated it as a string before a laterHSETorSADDtried to reuse the same name for a hash or set. - A typo or copy-pasted key name that happens to collide with an existing key of a different purpose.
- A key-naming convention without a type suffix, making it easy for two different parts of an application to disagree about what a given key should hold.
- Reusing a counter or flag key (typically a string, from
INCRorSET) for a different structure later in the same codebase.
How to fix it
- Check the key’s actual type before assuming what it holds:
TYPE mykey. It returnsstring,list,set,zset,hash,stream(orvectorseton Redis 8), ornoneif the key doesn’t exist. - If the key was created with the wrong type by mistake, delete it and recreate it with the intended type:
DEL mykeythen the correct command (HSET,SADD,RPUSH, etc.). - If two different features genuinely need two different structures, give them different key names — a naming convention that encodes the type (
user:1:profileas a hash,user:1:sessionsas a set) prevents this class of collision. - In application code, check
TYPE(or catch theWRONGTYPEerror explicitly) before an operation that depends on the key already existing with a specific type, rather than assuming it. - If this shows up after a deploy, check for a recent change to which command writes that key — a refactor that switched from
SETtoHSETwithout migrating existing keys is a common trigger.
In Tablyne
Tablyne’s key inspector shows a key’s type in its header as soon as you open it, and the keyspace sidebar can filter keys by type, so you can see that a key is a string rather than a hash before you run a command against it, instead of finding out from a WRONGTYPE error.
See every key’s real type before you touch it. Download Tablyne or open the Redis & Valkey GUI.
Sources
Frequently asked questions
Why does the same key name work in one command and not another? +
Every Redis key has exactly one data type (string, list, set, hash, sorted set, stream, and a few others) for as long as it exists. A command that expects a different type than what's actually stored — HGET on a string, LPUSH on a set — always fails with WRONGTYPE, regardless of the key's name or when it was created.
Does WRONGTYPE mean the key is corrupted? +
No. It's Redis correctly enforcing that a key has one fixed type. It usually means the application (or a script, or a colleague) wrote to that key with a different command than the one now reading it, or the key name collided with an existing key of another type.
How do I find out what type a key already is? +
Run TYPE <key>. It returns one of string, list, set, zset, hash, stream (or vectorset), or none if the key doesn't exist — checking it before an operation avoids the error entirely.