[cluster-mode] redis-cli show database number after using SELECT

[Grokzen]

When running in cluster mode, changing db is not supported but redis-cli still show the database number after SELECT is used.

This is a example output from redis-cli

redis 127.0.0.1:7000[2]> SELECT 0
OK
redis 127.0.0.1:7000> SELECT 1
(error) ERR SELECT is not allowed in cluster mode
redis 127.0.0.1:7000[1]> SELECT 2
(error) ERR SELECT is not allowed in cluster mode
redis 127.0.0.1:7000[2]> SELECT 3
(error) ERR SELECT is not allowed in cluster mode

[qwitwa]

This happens regardless of clustering settings if you use a large enough database number: Redis-cli simply doesn't check that the select didn't error when changing the db number it displays.

127.0.0.1:6379> SELECT 4
OK
127.0.0.1:6379[4]> SET foo bar
OK
127.0.0.1:6379[4]> SELECT 1000
(error) ERR invalid DB index
27.0.0.1:6379[1000]> SET foo confusingly
OK
27.0.0.1:6379[1000]> get foo
"confusingly"
127.0.0.1:6379[1000]> select 4
OK
127.0.0.1:6379[4]> get foo
"confusingly"

[HeartSaVioR]

Seems like it was fixed by unstable branch. Could you update & build and try again?

-- updated
@qwitwa's issue doesn't resolved. It seems resolved only display layer.

127.0.0.1:6379> del foo
(integer) 0
127.0.0.1:6379> select 4
OK
127.0.0.1:6379[4]> del foo
(integer) 1
127.0.0.1:6379[4]> set foo bar
OK
127.0.0.1:6379[4]> select 1000
(error) ERR invalid DB index
127.0.0.1:6379> get foo
"bar"
127.0.0.1:6379> select 0
OK
127.0.0.1:6379> get foo
(nil)

[HeartSaVioR]

Maybe I've fixed this issue by #2283, could you please check and verify? Thanks!

[qwitwa]

After building from the fix-redis-cli-select-db branch:

127.0.0.1:6379> get foo
(nil)
127.0.0.1:6379> select 2
OK
127.0.0.1:6379[2]> get foo
"baz"
127.0.0.1:6379[2]> select 1000
(error) ERR invalid DB index
127.0.0.1:6379[2]> get foo
"baz"
127.0.0.1:6379[2]>

Yup, looks good.

[qwitwa]

\o/ All tests passed without errors!

Tests pass.

[HeartSaVioR]

@qwitwa Thanks for verifying!

[mattsta]

Closing because tracking in #2283