When getKeysUsingKeySpecs processes a command with more than one key-spec,
and called with a total of more than 256 keys, it'll call getKeysPrepareResult again,
but since numkeys isn't updated, getKeysPrepareResult will not bother to copy key
names from the old result (leaving these slots uninitialized). Furthermore, it did not
consider the keys it already found when allocating more space.
Co-authored-by: Oran Agra <oran@redislabs.com>
The negative offset check was added in #9052, we realized
that this is a non-mandatory breaking change and we would
like to add it only in 8.0.
This reverts PR #9052, will be re-introduced later in 8.0.
Optimized the performance of the SCAN command in a few ways:
1. Move the key filtering (by MATCH pattern) in the scan callback,
so as to avoid collecting them for later filtering.
2. Reduce a many memory allocations and copying (use a reference
to the original sds, instead of creating an robj, an excessive 2 mallocs
and one string duplication)
3. Compare TYPE filter directly (as integers), instead of inefficient string
compare per key.
4. fixed a small bug: when scan zset and hash types, maxiterations uses
a more accurate number to avoid wrong double maxiterations.
Changes **postponed** for a later version (8.0):
1. Prepare to move the TYPE filtering to the scan callback as well. this was
put on hold since it has side effects that can be considered a breaking
change, which is that we will not attempt to do lazy expire (delete) a key
that was filtered by not matching the TYPE (changing it would mean TYPE filter
starts behaving the same as MATCH filter already does in that respect).
2. when the specified key TYPE filter is an unknown type, server will reply a error
immediately instead of doing a full scan that comes back empty handed.
Benchmark result:
For different scenarios, we obtained about 30% or more performance improvement.
Co-authored-by: Oran Agra <oran@redislabs.com>
Originally, when "tls-cluster" is enabled, `port` is set to TLS port. In order to support non-TLS clients, `pport` is used to propagate TCP port across cluster nodes. However when "tls-cluster" is disabled, `port` is set to TCP port, and `pport` is not used, which means the cluster cannot provide TLS service unless "tls-cluster" is on.
```
typedef struct {
// ...
uint16_t port; /* Latest known clients port (TLS or plain). */
uint16_t pport; /* Latest known clients plaintext port. Only used if the main clients port is for TLS. */
// ...
} clusterNode;
```
```
typedef struct {
// ...
uint16_t port; /* TCP base port number. */
uint16_t pport; /* Sender TCP plaintext port, if base port is TLS */
// ...
} clusterMsg;
```
This PR renames `port` and `pport` in `clusterNode` to `tcp_port` and `tls_port`, to record both ports no matter "tls-cluster" is enabled or disabled.
This allows to provide TLS service to clients when "tls-cluster" is disabled: when displaying cluster topology, or giving `MOVED` error, server can provide TLS or TCP port according to client's connection type, no matter what type of connection cluster bus is using.
For backwards compatibility, `port` and `pport` in `clusterMsg` are preserved, when "tls-cluster" is enabled, `port` is set to TLS port and `pport` is set to TCP port, when "tls-cluster" is disabled, `port` is set to TCP port and `pport` is set to TLS port (instead of 0).
Also, in the nodes.conf file, a new aux field displaying an extra port is added to complete the persisted info. We may have `tls_port=xxxxx` or `tcp_port=xxxxx` in the aux field, to complete the cluster topology, while the other port is stored in the normal `<ip>:<port>` field. The format is shown below.
```
<node-id> <ip>:<tcp_port>@<cport>,<hostname>,shard-id=...,tls-port=6379 myself,master - 0 0 0 connected 0-1000
```
Or we can switch the position of two ports, both can be correctly resolved.
```
<node-id> <ip>:<tls_port>@<cport>,<hostname>,shard-id=...,tcp-port=6379 myself,master - 0 0 0 connected 0-1000
```
blocking RM_Call was introduced on: #11568, It allows a module to perform
blocking commands and get the reply asynchronously.If the command gets
block, a special promise CallReply is returned that allow to set the unblock
handler. The unblock handler will be called when the command invocation
finish and it gets, as input, the command real reply.
The issue was that the real CallReply was created using a stack allocated
RedisModuleCtx which is no longer available after the unblock handler finishes.
So if the module keeps the CallReply after the unblock handler finished, the
CallReply holds a pointer to invalid memory and will try to access it when the
CallReply will be released.
The solution is to create the CallReply with a NULL context to make it totally
detached and can be freed freely when the module wants.
Test was added to cover this case, running the test with valgrind before the
fix shows the use after free error. With the fix, there are no valgrind errors.
unrelated: adding a missing `$rd close` in many tests in that file.
Apart from adding the missing coverage, this PR also adds `blockedBeforeSleep`
that gathers all block-related functions from `beforeSleep`
The order inside `blockedBeforeSleep` is different: now `handleClientsBlockedOnKeys`
(which may unblock clients) is called before `processUnblockedClients` (which handles
unblocked clients).
It makes sense to have this order.
There are no visible effects of the wrong ordering, except some cleanups of the now-unblocked
client would have happen in the next `beforeSleep` (will now happen in the current one)
The reason we even got into it is because i triggers an assertion in logresreq.c (breaking
the assumption that `unblockClient` is called **before** actually flushing the reply to the socket):
`handleClientsBlockedOnKeys` is called, then it calls `moduleUnblockClientOnKey`, which calls
`moduleUnblockClient`, which adds the client to `moduleUnblockedClients` back to `beforeSleep`,
we call `handleClientsWithPendingWritesUsingThreads`, it writes the data of buf to the client, so
`client->bufpos` became 0
On the next `beforeSleep`, we call `moduleHandleBlockedClients`, which calls `unblockClient`,
which calls `reqresAppendResponse`, triggering the assert. (because the `bufpos` is 0) - see https://github.com/redis/redis/pull/12301#discussion_r1226386716
To determine when everything was stable, we couldn't just query the nodename since they aren't API visible by design. Instead, we were using a proxy piece of information which was bumping the epoch and waiting for everyone to observe that. This works for making source Node 0 and Node 1 had pinged, and Node 0 and Node 2 had pinged, but did not guarantee that Node 1 and Node 2 had pinged. Although unlikely, this can cause this failure message. To fix it I hijacked hostnames and used its validation that it has been propagated, since we know that it is stable.
I also noticed while stress testing this sometimes the test took almost 4.5 seconds to finish, which is really close to the current 5 second limit of the log check, so I bumped that up as well just to make it a bit more consistent.
Introduced by https://github.com/redis/redis/pull/11923 (Redis 7.2 RC2)
It's very weird and counterintuitive that `RM_ReplyWithError` requires the error-code
**without** a hyphen while `RM_ReplyWithErrorFormat` requires either the error-code
**with** a hyphen or no error-code at all
```
RedisModule_ReplyWithError(ctx, "BLA bla bla");
```
vs.
```
RedisModule_ReplyWithErrorFormat(ctx, "-BLA %s", "bla bla");
```
This commit aligns RM_ReplyWithErrorFormat to behvae like RM_ReplyWithError.
it's a breaking changes but it's done before 7.2 goes GA.
When a connection that's subscribe to a channel emits PUBLISH inside MULTI-EXEC,
the push notification messes up the EXEC response.
e.g. MULTI, PING, PUSH foo bar, PING, EXEC
the EXEC's response will contain: PONG, {message foo bar}, 1. and the second PONG
will be delivered outside the EXEC's response.
Additionally, this PR changes the order of responses in case of a plain PUBLISH (when
the current client also subscribed to it), by delivering the push after the command's
response instead of before it.
This also affects modules calling RM_PublishMessage in a similar way, so that we don't
run the risk of getting that push mixed together with the module command's response.
Now we will check the offset in zrangeGenericCommand.
With a negative offset, we will throw an error and return.
This also resolve the issue of zeroing the destination key
in case of the "store" variant when we input a negative offset.
```
127.0.0.1:6379> set key value
OK
127.0.0.1:6379> zrangestore key myzset 0 10 byscore limit -1 10
(integer) 0
127.0.0.1:6379> exists key
(integer) 0
```
This change affects the following commands:
- ZRANGE / ZRANGESTORE / ZRANGEBYLEX / ZRANGEBYSCORE
- ZREVRANGE / ZREVRANGEBYSCORE / ZREVRANGEBYLEX
For geosearch and georadius we have already test coverage for wrong type, but we dont have for geodist, geohash, geopos commands. So adding the wrong type test cases for geodist, geohash, geopos commands.
Existing code, we have verify_geo_edge_response_bymember function for wrong type test cases which has member as an option. But the function is being called in other test cases where the output is not inline with these commnds(geodist, geohash, geopos). So I could not include these commands(geodist, geohash, geopos) as part of existing function, hence implemented a new function verify_geo_edge_response_generic and called from the test case.
Observed that the sanitizer reported memory leak as clean up is not done
before the process termination in negative/following cases:
**- when we passed '--invalid' as option to redis-server.**
```
-vm:~/mem-leak-issue/redis$ ./src/redis-server --invalid
*** FATAL CONFIG FILE ERROR (Redis 255.255.255) ***
Reading the configuration file, at line 2
>>> 'invalid'
Bad directive or wrong number of arguments
=================================================================
==865778==ERROR: LeakSanitizer: detected memory leaks
Direct leak of 8 byte(s) in 1 object(s) allocated from:
#0 0x7f0985f65867 in __interceptor_malloc ../../../../src/libsanitizer/asan/asan_malloc_linux.cpp:145
#1 0x558ec86686ec in ztrymalloc_usable_internal /home/ubuntu/mem-leak-issue/redis/src/zmalloc.c:117
#2 0x558ec86686ec in ztrymalloc_usable /home/ubuntu/mem-leak-issue/redis/src/zmalloc.c:135
#3 0x558ec86686ec in ztryrealloc_usable_internal /home/ubuntu/mem-leak-issue/redis/src/zmalloc.c:276
#4 0x558ec86686ec in zrealloc /home/ubuntu/mem-leak-issue/redis/src/zmalloc.c:327
#5 0x558ec865dd7e in sdssplitargs /home/ubuntu/mem-leak-issue/redis/src/sds.c:1172
#6 0x558ec87a1be7 in loadServerConfigFromString /home/ubuntu/mem-leak-issue/redis/src/config.c:472
#7 0x558ec87a13b3 in loadServerConfig /home/ubuntu/mem-leak-issue/redis/src/config.c:718
#8 0x558ec85e6f15 in main /home/ubuntu/mem-leak-issue/redis/src/server.c:7258
#9 0x7f09856e5d8f in __libc_start_call_main ../sysdeps/nptl/libc_start_call_main.h:58
SUMMARY: AddressSanitizer: 8 byte(s) leaked in 1 allocation(s).
```
**- when we pass '--port' as option and missed to add port number to redis-server.**
```
vm:~/mem-leak-issue/redis$ ./src/redis-server --port
*** FATAL CONFIG FILE ERROR (Redis 255.255.255) ***
Reading the configuration file, at line 2
>>> 'port'
wrong number of arguments
=================================================================
==865846==ERROR: LeakSanitizer: detected memory leaks
Direct leak of 8 byte(s) in 1 object(s) allocated from:
#0 0x7fdcdbb1f867 in __interceptor_malloc ../../../../src/libsanitizer/asan/asan_malloc_linux.cpp:145
#1 0x557e8b04f6ec in ztrymalloc_usable_internal /home/ubuntu/mem-leak-issue/redis/src/zmalloc.c:117
#2 0x557e8b04f6ec in ztrymalloc_usable /home/ubuntu/mem-leak-issue/redis/src/zmalloc.c:135
#3 0x557e8b04f6ec in ztryrealloc_usable_internal /home/ubuntu/mem-leak-issue/redis/src/zmalloc.c:276
#4 0x557e8b04f6ec in zrealloc /home/ubuntu/mem-leak-issue/redis/src/zmalloc.c:327
#5 0x557e8b044d7e in sdssplitargs /home/ubuntu/mem-leak-issue/redis/src/sds.c:1172
#6 0x557e8b188be7 in loadServerConfigFromString /home/ubuntu/mem-leak-issue/redis/src/config.c:472
#7 0x557e8b1883b3 in loadServerConfig /home/ubuntu/mem-leak-issue/redis/src/config.c:718
#8 0x557e8afcdf15 in main /home/ubuntu/mem-leak-issue/redis/src/server.c:7258
#9 0x7fdcdb29fd8f in __libc_start_call_main ../sysdeps/nptl/libc_start_call_main.h:58
Indirect leak of 10 byte(s) in 1 object(s) allocated from:
#0 0x7fdcdbb1fc18 in __interceptor_realloc ../../../../src/libsanitizer/asan/asan_malloc_linux.cpp:164
#1 0x557e8b04f9aa in ztryrealloc_usable_internal /home/ubuntu/mem-leak-issue/redis/src/zmalloc.c:287
#2 0x557e8b04f9aa in ztryrealloc_usable /home/ubuntu/mem-leak-issue/redis/src/zmalloc.c:317
#3 0x557e8b04f9aa in zrealloc_usable /home/ubuntu/mem-leak-issue/redis/src/zmalloc.c:342
#4 0x557e8b033f90 in _sdsMakeRoomFor /home/ubuntu/mem-leak-issue/redis/src/sds.c:271
#5 0x557e8b033f90 in sdsMakeRoomFor /home/ubuntu/mem-leak-issue/redis/src/sds.c:295
#6 0x557e8b033f90 in sdscatlen /home/ubuntu/mem-leak-issue/redis/src/sds.c:486
#7 0x557e8b044e1f in sdssplitargs /home/ubuntu/mem-leak-issue/redis/src/sds.c:1165
#8 0x557e8b188be7 in loadServerConfigFromString /home/ubuntu/mem-leak-issue/redis/src/config.c:472
#9 0x557e8b1883b3 in loadServerConfig /home/ubuntu/mem-leak-issue/redis/src/config.c:718
#10 0x557e8afcdf15 in main /home/ubuntu/mem-leak-issue/redis/src/server.c:7258
#11 0x7fdcdb29fd8f in __libc_start_call_main ../sysdeps/nptl/libc_start_call_main.h:58
SUMMARY: AddressSanitizer: 18 byte(s) leaked in 2 allocation(s).
```
As part analysis found that the sdsfreesplitres is not called when this condition checks are being hit.
Output after the fix:
```
vm:~/mem-leak-issue/redis$ ./src/redis-server --invalid
*** FATAL CONFIG FILE ERROR (Redis 255.255.255) ***
Reading the configuration file, at line 2
>>> 'invalid'
Bad directive or wrong number of arguments
vm:~/mem-leak-issue/redis$
===========================================
vm:~/mem-leak-issue/redis$ ./src/redis-server --jdhg
*** FATAL CONFIG FILE ERROR (Redis 255.255.255) ***
Reading the configuration file, at line 2
>>> 'jdhg'
Bad directive or wrong number of arguments
---------------------------------------------------------------------------
vm:~/mem-leak-issue/redis$ ./src/redis-server --port
*** FATAL CONFIG FILE ERROR (Redis 255.255.255) ***
Reading the configuration file, at line 2
>>> 'port'
wrong number of arguments
```
Co-authored-by: Oran Agra <oran@redislabs.com>
Adds API
- RedisModule_CommandFilterGetClientId()
Includes addition to commandfilter test module to validate that it works
by performing the same command from 2 different clients
This PR adds a human readable name to a node in clusters that are visible as part of error logs. This is useful so that admins and operators of Redis cluster have better visibility into failures without having to cross-reference the generated ID with some logical identifier (such as pod-ID or EC2 instance ID). This is mentioned in #8948. Specific nodenames can be set by using the variable cluster-announce-human-nodename. The nodename is gossiped using the clusterbus extension in #9530.
Co-authored-by: Madelyn Olson <madelyneolson@gmail.com>
## Issue:
When a dict has a long chain or the length of the chain is longer than
the number of samples, we will never be able to sample the elements
at the end of the chain using dictGetSomeKeys().
This could mean that SRANDMEMBER can be hang in and endless loop.
The most severe case, is the pathological case of when someone uses SCAN+DEL
or SSCAN+SREM creating an unevenly distributed dict.
This was amplified by the recent change in #11692 which prevented a
down-sizing rehashing while there is a fork.
## Solution
1. Before, we will stop sampling when we reach the maximum number
of samples, even if there is more data after the current chain.
Now when we reach the maximum we use the Reservoir Sampling
algorithm to fairly sample the end of the chain that cannot be sampled
2. Fix the rehashing code, so that the same as it allows rehashing for up-sizing
during fork when the ratio is extreme, it will allow it for down-sizing as well.
Issue was introduced (or became more severe) by #11692
Co-authored-by: Oran Agra <oran@redislabs.com>
In SPOP, when COUNT is greater than or equal to set's size,
we will remove the set. In dbDelete, we will do DEL or UNLINK
according to the lazy flag. This is also required for propagate.
In RESTORE, we won't store expired keys into the db, see #7472.
When used together with REPLACE, it should emit a DEL or UNLINK
according to the lazy flag.
This PR also adds tests to cover the propagation. The RESTORE
test will also cover #7472.
* Add command being unblocked cause another command to get unblocked execution order test
In #12301, we observed that if the
`while(listLength(server.ready_keys) != 0)`
in handleClientsBlockedOnKeys is changed to
`if(listLength(server.ready_keys) != 0)`,
the order of command execution will change.
It is wrong to change that. It means that if a command
being unblocked causes another command to get unblocked
(like a BLMOVE would do), then the new unblocked command
will wait for later to get processed rather than right away.
It'll not have any real implication if we change that since
we do call handleClientsBlockedOnKeys in beforeSleep again,
and redis will still behave correctly, but we don't change that.
An example:
1. $rd1 blmove src{t} dst{t} left right 0
2. $rd2 blmove dst{t} src{t} right left 0
3. $rd3 set key1{t}, $rd3 lpush src{t}, $rd3 set key2{t} in a pipeline
The correct order would be:
1. set key1{t}
2. lpush src{t}
3. lmove src{t} dst{t} left right
4. lmove dst{t} src{t} right left
5. set key2{t}
The wrong order would be:
1. set key1{t}
2. lpush src{t}
3. lmove src{t} dst{t} left right
4. set key2{t}
5. lmove dst{t} src{t} right left
This PR adds corresponding test to cover it.
* Add comment near while(listLength(server.ready_keys) != 0)
For the XREADGROUP BLOCK > scenario, there is an endless loop.
Due to #11012, it keep going, reprocess command -> blockForKeys -> reprocess command
The right fix is to avoid an endless loop in handleClientsBlockedOnKey and handleClientsBlockedOnKeys,
looks like there was some attempt in handleClientsBlockedOnKeys but maybe not sufficiently good,
and it looks like using a similar trick in handleClientsBlockedOnKey is complicated.
i.e. stashing the list on the stack and iterating on it after creating a fresh one for future use,
is problematic since the code keeps accessing the global list.
Co-authored-by: Oran Agra <oran@redislabs.com>
This will increase the size of an already large COB (one already passed
the threshold for disconnection)
This could also mean that we'll attempt to write that data to the socket
and the replica will manage to read it, which will result in an
undesired partial sync (undesired for the test)
In 7.2, After 971b177fa we make sure (assert) that
the duration has been recorded when resetting the client.
This is not true for rejected commands.
The use case I found is a blocking command that an ACL rule changed before
it was unblocked, and while reprocessing it, the command rejected and triggered the assert.
The PR reset the command duration inside rejectCommand / rejectCommandSds.
Co-authored-by: Oran Agra <oran@redislabs.com>
In #11963, some new tests about eventloop duration were added, which includes time measurement in TCL scripts. This has caused some unexpected CI failures, such as #12169 and #12177, due to slow test servers or some performance jittering.
Added missing test case coverage for below scenarios:
1. The command only works if all the specified slots are, from
the point of view of the node receiving the command, currently
not assigned. A node will refuse to take ownership for slots that
already belong to some other node (including itself).
2. The command fails if the same slot is specified multiple times.
So far clients being blocked and unblocked by a module command would
update the c->woff variable and so WAIT was ineffective and got released
without waiting for the command actions to propagate.
This seems to have existed since forever, but not for RM_BlockClientOnKeys.
It is problematic though to know if the module did or didn't propagate
anything in that command, so for now, instead of adding an API, we'll
just update the woff to the latest offset when unblocking, this will
cause the client to possibly wait excessively, but that's not that bad.
XREAD only supports a special ID of $ and XREADGROUP only supports ^.
make sure not to suggest the wrong one when rerunning an error about unbalanced ID arguments
Co-authored-by: Oran Agra <oran@redislabs.com>
This pr can get two performance benefits:
1. Stop redundant initialization when most robj objects are created
2. LRU_CLOCK will no longer be called in io threads, so we can avoid the `atomicGet`
Another code optimization:
deleted the redundant judgment in dbSetValue, no matter in LFU or LRU, the lru field inold
robj is always the freshest (it is always updated in lookupkey), so we don't need to judge if in LFU
This commit excludes aux fields from the output of the `cluster nodes` and `cluster replicas` command.
We may decide to re-introduce them in some form or another in the future, but not in v7.2.
For zsets that will eventually be stored as the skiplist encoding (has a dict),
we can convert it to skiplist ahead of time. This change checks the number
of arguments in the ZADD command, and converts the data-structure
if the number of new entries exceeds the listpack-max-entries configuration.
This can cause us to over-allocate memory if there are duplicate entries in the
input, which is unexpected.
For ZRANGESTORE, we know the size of the zset, so we can expand
the dict in advance, to avoid the temporary dict from being rehashed
while it grows.
Simple benchmarks shows it provides some 4% improvement in ZADD and 20% in ZRANGESTORE
Current tests for BITFIELD_RO command are skipped in the external mode,
and therefore reply-schemas-validator reports a coverage error.
This PR adds basic tests to increase coverage.
The measured latency(duration) includes the list below, which can be shown by `INFO STATS`.
```
eventloop_cycles // ever increasing counter
eventloop_duration_sum // cumulative duration of eventloop in microseconds
eventloop_duration_cmd_sum // cumulative duration of executing commands in microseconds
instantaneous_eventloop_cycles_per_sec // average eventloop count per second in recent 1.6s
instantaneous_eventloop_duration_usec // average single eventloop duration in recent 1.6s
```
Also added some experimental metrics, which are shown only when `INFO DEBUG` is called.
This section isn't included in the default INFO, or even in `INFO ALL` and the fields in this section
can change in the future without considering backwards compatibility.
```
eventloop_duration_aof_sum // cumulative duration of writing AOF
eventloop_duration_cron_sum // cumulative duration cron jobs (serverCron, beforeSleep excluding IO and AOF)
eventloop_cmd_per_cycle_max // max number of commands executed in one eventloop
eventloop_duration_max // max duration of one eventloop
```
All of these are being reset by CONFIG RESETSTAT
The test failed on MacOS:
```
*** [err]: EXPIRE precision is now the millisecond in tests/unit/expire.tcl
Expected 'somevalue {}' to equal or match '{} {}'
```
`set a [r get x]`, even though we tried 10 times, sometimes we
still get {}, this is a time-sensitive test.
In this PR, we add the following changes:
1. More attempts, change it from 10 to 30.
2. More tolerant, change the `after 900` to `after 800`.
In addition, we judging $a in advance and changing `after 1100`
to `after 300`, this will save us some times.
When `RM_ZsetAdd()`/`RM_ZsetIncrby()`/`RM_StreamAdd()` fails, if a new key happens to
be created using `moduleCreateEmptyKey()`, we should clean up the empty key.
## Test
1) Add new module commands(`zset.add` and `zset.incrby`) to cover `RM_ZsetAdd()`/`RM_ZsetIncrby()`.
2) Add a large-memory test to cover `RM_StreamAdd()`.
Tests occasionally fail since #12000:
```
*** [err]: query buffer resized correctly when not idle in tests/unit/querybuf.tcl
Expected 0 > 32768 (context: type eval line 11 cmd {assert {$orig_test_client_qbuf > 32768}} proc ::test)
*** [err]: query buffer resized correctly with fat argv in tests/unit/querybuf.tcl
query buffer should not be resized when client idle time smaller than 2s
```
The reason may be because we set hz to 100, querybuf shrinks before we count
client_query_buffer. We avoid this problem by setting pause-cron to 1.
1. reset the readraw mode after a test that uses it. undetected since the
only test after that on the same server didn't read any replies.
2. fix a cross slot issue that was undetected in cluster mode because
readraw doesn't throw exceptions on errors.
Minor test case addition for DECR and DECRBY.
Currently DECR and DECRBY do not have test case coverage for the
scenarios where they run on a non-existing key.
In order to speed up tests, avoid saving an RDB (mostly notable on shutdown),
except for tests that explicitly test the RDB mechanism
In addition, use `shutdown-on-sigterm force` to prevetn shutdown from failing
in case the server is in the middle of the initial AOFRW
Also a a test that checks that the `shutdown-on-sigterm default` is to refuse
shutdown if there's an initial AOFRW
Co-authored-by: Guy Benoish <guy.benoish@redislabs.com>
Minor test case addition.
Currently GETRANGE command does not have the test case coverage for the scenarios:
An error is returned when key exists but of different type
Added missing test cases for getrange command.
this pr fix two wrongs：
1. When client’s querybuf is pre-allocated for a fat argv, we need to update the
querybuf_peak of the client immediately to completely avoid the unexpected
shrinking of querybuf in the next clientCron (before data arrives to set the peak).
2. the protocol's bulklen does not include `\r\n`, but the allocation and the data we
read does. so in `clientsCronResizeQueryBuffer`, the `resize` or `querybuf_peak`
should add these 2 bytes.
the first bug is likely to hit us on large payloads over slow connections, in which case
transferring the payload can take longer and a cron event will be triggered (specifically
if there are not a lot of clients)
There is are some missing test cases for SUBSTR command.
These might already be covered by GETRANGE, but no harm in adding them since they're simple.
Added 3 test case.
* start > stop
* start and stop both greater than string length
* when no key is present.
Currently LINSERT command does not have the test case coverage for following scenarios.
1. When key does not exist, it is considered an empty list and no operation is performed.
2. An error is returned when key exists but does not hold a list value.
Added above two missing test cases for linsert command.
* Add RM_ReplyWithErrorFormat that can support format
Reply with the error create from a printf format and arguments.
If the error code is already passed in the string 'fmt', the error
code provided is used, otherwise the string "-ERR " for the generic
error code is automatically added.
The usage is, for example:
RedisModule_ReplyWithErrorFormat(ctx, "An error: %s", "foo");
RedisModule_ReplyWithErrorFormat(ctx, "-WRONGTYPE Wrong Type: %s", "foo");
The function always returns REDISMODULE_OK.
The MacOS CI in github actions often hangs without any logs. GH argues that
it's due to resource utilization, either running out of disk space, memory, or CPU
starvation, and thus the runner is terminated.
This PR contains multiple attempts to resolve this:
1. introducing pause_process instead of SIGSTOP, which waits for the process
to stop before resuming the test, possibly resolving race conditions in some tests,
this was a suspect since there was one test that could result in an infinite loop in that
case, in practice this didn't help, but still a good idea to keep.
2. disable the `save` config in many tests that don't need it, specifically ones that use
heavy writes and could create large files.
3. change the `populate` proc to use short pipeline rather than an infinite one.
4. use `--clients 1` in the macos CI so that we don't risk running multiple resource
demanding tests in parallel.
5. enable `--verbose` to be repeated to elevate verbosity and print more info to stdout
when a test or a server starts.
We do have ZREMRANGEBYLEX tests, but it is a stress test
marked with slow tag and then skipped in reply-schemas daily.
In the past, we were able to succeed on a daily, i guess
it was because there were some random command executions,
such as corrupt-dump-fuzzy, which might call it.
These test examples are taken from ZRANGEBYLEX basics test.
Add `RM_RdbLoad()` and `RM_RdbSave()` to load/save RDB files from the module API.
In our use case, we have our clustering implementation as a module. As part of this
implementation, the module needs to trigger RDB save operation at specific points.
Also, this module delivers RDB files to other nodes (not using Redis' replication).
When a node receives an RDB file, it should be able to load the RDB. Currently,
there is no module API to save/load RDB files.
This PR adds four new APIs:
```c
RedisModuleRdbStream *RM_RdbStreamCreateFromFile(const char *filename);
void RM_RdbStreamFree(RedisModuleRdbStream *stream);
int RM_RdbLoad(RedisModuleCtx *ctx, RedisModuleRdbStream *stream, int flags);
int RM_RdbSave(RedisModuleCtx *ctx, RedisModuleRdbStream *stream, int flags);
```
The first step is to create a `RedisModuleRdbStream` object. This PR provides a function to
create RedisModuleRdbStream from the filename. (You can load/save RDB with the filename).
In the future, this API can be extended if needed:
e.g., `RM_RdbStreamCreateFromFd()`, `RM_RdbStreamCreateFromSocket()` to save/load
RDB from an `fd` or a `socket`.
Usage:
```c
/* Save RDB */
RedisModuleRdbStream *stream = RedisModule_RdbStreamCreateFromFile("example.rdb");
RedisModule_RdbSave(ctx, stream, 0);
RedisModule_RdbStreamFree(stream);
/* Load RDB */
RedisModuleRdbStream *stream = RedisModule_RdbStreamCreateFromFile("example.rdb");
RedisModule_RdbLoad(ctx, stream, 0);
RedisModule_RdbStreamFree(stream);
```
The existing logic for killing pub-sub clients did not handle the `allchannels`
permission correctly. For example, if you:
ACL SETUSER foo allchannels
Have a client authenticate as the user `foo` and subscribe to a channel, and then:
ACL SETUSER foo resetchannels
The subscribed client would not be disconnected, though new clients under that user
would be blocked from subscribing to any channels.
This was caused by an incomplete optimization in `ACLKillPubsubClientsIfNeeded`
checking whether the new channel permissions were a strict superset of the old ones.
This PR fix several unrelated bugs that were discovered by the same set of tests
(WAITAOF tests in #11713), could make the `WAITAOF` test hang.
The change in `backgroundRewriteDoneHandler` is about MP-AOF.
That leftover / old code assumes that we started a new AOF file just now
(when we have a new base into which we're gonna incrementally write), but
the fact is that with MP-AOF, the fork done handler doesn't really affect the
incremental file being maintained by the parent process, there's no reason to
re-issue `SELECT`, and no reason to update any of the fsync variables in that flow.
This should have been deleted with MP-AOF (introduced in #9788, 7.0).
The damage is that the update to `aof_fsync_offset` will cause us to miss an fsync
in `flushAppendOnlyFile`, that happens if we stop write commands in `AOF_FSYNC_EVERYSEC`
while an AOFRW is in progress. This caused a new `WAITAOF` test to sometime hang forever.
Also because of MP-AOF, we needed to change `aof_fsync_offset` to `aof_last_incr_fsync_offset`
and match it to `aof_last_incr_size` in `flushAppendOnlyFile`. This is because in the past we compared
`aof_fsync_offset` and `aof_current_size`, but with MP-AOF it could be the total AOF file will be
smaller after AOFRW, and the (already existing) incr file still has data that needs to be fsynced.
The change in `flushAppendOnlyFile`, about the `AOF_FSYNC_ALWAYS`, it is follow #6053
(the details is in #5985), we also check `AOF_FSYNC_ALWAYS` to handle a case where
appendfsync is changed from everysec to always while there is data that's written but not yet fsynced.
Since we remove the COMMAND COUNT call in sentinel test in #11950,
reply-schemas-validator started reporting this error:
```
WARNING! The following commands were not hit at all:
command|count
ERROR! at least one command was not hit by the tests
```
This PR add a COMMAND COUNT test to cover it and also fix some
typos in req-res-log-validator.py
This PR allows clients to send information about the client library to redis
to be displayed in CLIENT LIST and CLIENT INFO.
Currently supports:
`CLIENT [lib-name | lib-ver] <value>`
Client libraries are expected to pipeline these right after AUTH, and ignore
the failure in case they're talking to an older version of redis.
These will be shown in CLIENT LIST and CLIENT INFO as:
* `lib-name` - meant to hold the client library name.
* `lib-ver` - meant to hold the client library version.
The values cannot contain spaces, newlines and any wild ASCII characters,
but all other normal chars are accepted, e.g `.`, `=` etc (same as CLIENT NAME).
The RESET command does NOT clear these, but they can be cleared to the
default by sending a command with a blank string.
Co-authored-by: Oran Agra <oran@redislabs.com>
This allows modules to register commands to existing ACL categories and blocks the creation of [sub]commands, datatypes and registering the configs outside of the OnLoad function.
For allowing modules to register commands to existing ACL categories,
This PR implements a new API int RM_SetCommandACLCategories() which takes a pointer to a RedisModuleCommand and a C string aclflags containing the set of space separated ACL categories.
Example, 'write slow' marks the command as part of the write and slow ACL categories.
The C string aclflags is tokenized by implementing a helper function categoryFlagsFromString(). Theses tokens are matched and the corresponding ACL categories flags are set by a helper function matchAclCategoriesFlags. The helper function categoryFlagsFromString() returns the corresponding categories_flags or returns -1 if some token not processed correctly.
If the module contains commands which are registered to existing ACL categories, the number of [sub]commands are tracked by num_commands_with_acl_categories in struct RedisModule. Further, the allowed command bit-map of the existing users are recomputed from the command_rules list, by implementing a function called ACLRecomputeCommandBitsFromCommandRulesAllUsers() for the existing users to have access to the module commands on runtime.
## Breaking change
This change requires that registering commands and subcommands only occur during a modules "OnLoad" function, in order to allow efficient recompilation of ACL bits. We also chose to block registering configs and types, since we believe it's only valid for those to be created during onLoad. We check for this onload flag in struct RedisModule to check if the call is made from the OnLoad function.
Co-authored-by: Madelyn Olson <madelyneolson@gmail.com>
I saw this error once, in the FreeBSD Daily CI:
```
*** [err]: Temp rdb will be deleted if we use bg_unlink when shutdown in tests/unit/shutdown.tcl
Expected [file exists /xxx/temp-10336.rdb] (context: type eval line 15 cmd {assert {[file exists $temp_rdb]}} proc ::test)
```
The log shows that bgsave was executed, and it was successfully executed in the end:
```
Starting test Temp rdb will be deleted if we use bg_unlink when shutdown in tests/unit/shutdown.tcl
10251:M 22 Feb 2023 11:37:25.441 * Background saving started by pid 10336
10336:C 22 Feb 2023 11:37:27.949 * DB saved on disk
10336:C 22 Feb 2023 11:37:27.949 * Fork CoW for RDB: current 0 MB, peak 0 MB, average 0 MB
10251:M 22 Feb 2023 11:37:28.060 * Background saving terminated with success
```
There may be two reasons:
1. The child process has been created, but it has not created
the temp rdb file yet, so [file exists $temp_rdb] check failed.
2. The child process bgsave has been executed successfully and the
temp file has been deleted, so [file exists $temp_rdb] check failed.
From the logs pint, it should be the case 2, case 1 is too extreme,
set rdb-key-save-delay to a higher value to ensure bgsave does not
succeed early to avoid this case.
Previously we would run the module command filters even upon blocked
command reprocessing. This could modify the command, and it's args.
This is irrelevant in the context of a command being reprocessed (it already
went through the filters), as well as breaks the crashed command lookup
that exists in the case of a reprocessed command.
fixes#11894.
Co-authored-by: Oran Agra <oran@redislabs.com>
Allow running blocking commands from within a module using `RM_Call`.
Today, when `RM_Call` is used, the fake client that is used to run command
is marked with `CLIENT_DENY_BLOCKING` flag. This flag tells the command
that it is not allowed to block the client and in case it needs to block, it must
fallback to some alternative (either return error or perform some default behavior).
For example, `BLPOP` fallback to simple `LPOP` if it is not allowed to block.
All the commands must respect the `CLIENT_DENY_BLOCKING` flag (including
module commands). When the command invocation finished, Redis asserts that
the client was not blocked.
This PR introduces the ability to call blocking command using `RM_Call` by
passing a callback that will be called when the client will get unblocked.
In order to do that, the user must explicitly say that he allow to perform blocking
command by passing a new format specifier argument, `K`, to the `RM_Call`
function. This new flag will tell Redis that it is allow to run blocking command
and block the client. In case the command got blocked, Redis will return a new
type of call reply (`REDISMODULE_REPLY_PROMISE`). This call reply indicates
that the command got blocked and the user can set the on_unblocked handler using
`RM_CallReplyPromiseSetUnblockHandler`.
When clients gets unblocked, it eventually reaches `processUnblockedClients` function.
This is where we check if the client is a fake module client and if it is, we call the unblock
callback instead of performing the usual unblock operations.
**Notice**: `RM_CallReplyPromiseSetUnblockHandler` must be called atomically
along side the command invocation (without releasing the Redis lock in between).
In addition, unlike other CallReply types, the promise call reply must be released
by the module when the Redis GIL is acquired.
The module can abort the execution on the blocking command (if it was not yet
executed) using `RM_CallReplyPromiseAbort`. the API will return `REDISMODULE_OK`
on success and `REDISMODULE_ERR` if the operation is already executed.
**Notice** that in case of misbehave module, Abort might finished successfully but the
operation will not really be aborted. This can only happened if the module do not respect
the disconnect callback of the blocked client.
For pure Redis commands this can not happened.
### Atomicity Guarantees
The API promise that the unblock handler will run atomically as an execution unit.
This means that all the operation performed on the unblock handler will be wrapped
with a multi exec transaction when replicated to the replica and AOF.
The API **do not** grantee any other atomicity properties such as when the unblock
handler will be called. This gives us the flexibility to strengthen the grantees (or not)
in the future if we will decide that we need a better guarantees.
That said, the implementation **does** provide a better guarantees when performing
pure Redis blocking command like `BLPOP`. In this case the unblock handler will run
atomically with the operation that got unblocked (for example, in case of `BLPOP`, the
unblock handler will run atomically with the `LPOP` operation that run when the command
got unblocked). This is an implementation detail that might be change in the future and the
module writer should not count on that.
### Calling blocking commands while running on script mode (`S`)
`RM_Call` script mode (`S`) was introduced on #0372. It is used for usecases where the
command that was invoked on `RM_Call` comes from a user input and we want to make
sure the user will not run dangerous commands like `shutdown`. Some command, such
as `BLPOP`, are marked with `NO_SCRIPT` flag, which means they will not be allowed on
script mode. Those commands are marked with `NO_SCRIPT` just because they are
blocking commands and not because they are dangerous. Now that we can run blocking
commands on RM_Call, there is no real reason not to allow such commands on script mode.
The underline problem is that the `NO_SCRIPT` flag is abused to also mark some of the
blocking commands (notice that those commands know not to block the client if it is not
allowed to do so, and have a fallback logic to such cases. So even if those commands
were not marked with `NO_SCRIPT` flag, it would not harm Redis, and today we can
already run those commands within multi exec).
In addition, not all blocking commands are marked with `NO_SCRIPT` flag, for example
`blmpop` are not marked and can run from within a script.
Those facts shows that there are some ambiguity about the meaning of the `NO_SCRIPT`
flag, and its not fully clear where it should be use.
The PR suggest that blocking commands should not be marked with `NO_SCRIPT` flag,
those commands should handle `CLIENT_DENY_BLOCKING` flag and only block when
it's safe (like they already does today). To achieve that, the PR removes the `NO_SCRIPT`
flag from the following commands:
* `blmove`
* `blpop`
* `brpop`
* `brpoplpush`
* `bzpopmax`
* `bzpopmin`
* `wait`
This might be considered a breaking change as now, on scripts, instead of getting
`command is not allowed from script` error, the user will get some fallback behavior
base on the command implementation. That said, the change matches the behavior
of scripts and multi exec with respect to those commands and allow running them on
`RM_Call` even when script mode is used.
### Additional RedisModule API and changes
* `RM_BlockClientSetPrivateData` - Set private data on the blocked client without the
need to unblock the client. This allows up to set the promise CallReply as the private
data of the blocked client and abort it if the client gets disconnected.
* `RM_BlockClientGetPrivateData` - Return the current private data set on a blocked client.
We need it so we will have access to this private data on the disconnect callback.
* On RM_Call, the returned reply will be added to the auto memory context only if auto
memory is enabled, this allows us to keep the call reply for longer time then the context
lifetime and does not force an unneeded borrow relationship between the CallReply and
the RedisModuleContext.
This change adds new module callbacks that can override the default password based authentication associated with ACLs. With this, Modules can register auth callbacks through which they can implement their own Authentication logic. When `AUTH` and `HELLO AUTH ...` commands are used, Module based authentication is attempted and then normal password based authentication is attempted if needed.
The new Module APIs added in this PR are - `RM_RegisterCustomAuthCallback` and `RM_BlockClientOnAuth` and `RedisModule_ACLAddLogEntryByUserName `.
Module based authentication will be attempted for all Redis users (created through the ACL SETUSER cmd or through Module APIs) even if the Redis user does not exist at the time of the command. This gives a chance for the Module to create the RedisModule user and then authenticate via the RedisModule API - from the custom auth callback.
For the AUTH command, we will support both variations - `AUTH <username> <password>` and `AUTH <password>`. In case of the `AUTH <password>` variation, the custom auth callbacks are triggered with “default” as the username and password as what is provided.
### RedisModule_RegisterCustomAuthCallback
```
void RM_RegisterCustomAuthCallback(RedisModuleCtx *ctx, RedisModuleCustomAuthCallback cb) {
```
This API registers a callback to execute to prior to normal password based authentication. Multiple callbacks can be registered across different modules. These callbacks are responsible for either handling the authentication, each authenticating the user or explicitly denying, or deferring it to other authentication mechanisms. Callbacks are triggered in the order they were registered. When a Module is unloaded, all the auth callbacks registered by it are unregistered. The callbacks are attempted, in the order of most recently registered callbacks, when the AUTH/HELLO (with AUTH field is provided) commands are called. The callbacks will be called with a module context along with a username and a password, and are expected to take one of the following actions:
(1) Authenticate - Use the RM_Authenticate* API successfully and return `REDISMODULE_AUTH_HANDLED`. This will immediately end the auth chain as successful and add the OK reply.
(2) Block a client on authentication - Use the `RM_BlockClientOnAuth` API and return `REDISMODULE_AUTH_HANDLED`. Here, the client will be blocked until the `RM_UnblockClient `API is used which will trigger the auth reply callback (provided earlier through the `RM_BlockClientOnAuth`). In this reply callback, the Module should authenticate, deny or skip handling authentication.
(3) Deny Authentication - Return `REDISMODULE_AUTH_HANDLED` without authenticating or blocking the client. Optionally, `err` can be set to a custom error message. This will immediately end the auth chain as unsuccessful and add the ERR reply.
(4) Skip handling Authentication - Return `REDISMODULE_AUTH_NOT_HANDLED` without blocking the client. This will allow the engine to attempt the next custom auth callback.
If none of the callbacks authenticate or deny auth, then password based auth is attempted and will authenticate or add failure logs and reply to the clients accordingly.
### RedisModule_BlockClientOnAuth
```
RedisModuleBlockedClient *RM_BlockClientOnAuth(RedisModuleCtx *ctx, RedisModuleCustomAuthCallback reply_callback,
void (*free_privdata)(RedisModuleCtx*,void*))
```
This API can only be used from a Module from the custom auth callback. If a client is not in the middle of custom module based authentication, ERROR is returned. Otherwise, the client is blocked and the `RedisModule_BlockedClient` is returned similar to the `RedisModule_BlockClient` API.
### RedisModule_ACLAddLogEntryByUserName
```
int RM_ACLAddLogEntryByUserName(RedisModuleCtx *ctx, RedisModuleString *username, RedisModuleString *object, RedisModuleACLLogEntryReason reason)
```
Adds a new entry in the ACL log with the `username` RedisModuleString provided. This simplifies the Module usage because now, developers do not need to create a Module User just to add an error ACL Log entry. Aside from accepting username (RedisModuleString) instead of a RedisModuleUser, it is the same as the existing `RedisModule_ACLAddLogEntry` API.
### Breaking changes
- HELLO command - Clients can now only set the client name and RESP protocol from the `HELLO` command if they are authenticated. Also, we now finish command arg validation first and return early with a ERR reply if any arg is invalid. This is to avoid mutating the client name / RESP from a command that would have failed on invalid arguments.
### Notable behaviors
- Module unblocking - Now, we will not allow Modules to block the client from inside the context of a reply callback (triggered from the Module unblock flow `moduleHandleBlockedClients`).
---------
Co-authored-by: Madelyn Olson <34459052+madolson@users.noreply.github.com>
There be a situation that satisfies WAIT, and then wrongly unblock
WAITAOF because we mix-use last_offset and last_numreplicas.
We update last_offset and last_numreplicas only when the condition
matches. i.e. output of either replicationCountAOFAcksByOffset or
replicationCountAcksByOffset is right.
In this case, we need to have separate last_ variables for each of
them. Added a last_aof_offset and last_aof_numreplicas for WAITAOF.
WAITAOF was added in #11713. Found while coding #11917.
A Test was added to validate that case.
WAITAOF wad added in #11713, its return is an array.
But forget to handle WAITAOF in last_offset and last_numreplicas,
causing WAITAOF to return a WAIT like reply.
Tests was added to validate that case (both WAIT and WAITAOF).
This PR also refactored processClientsWaitingReplicas a bit for better
maintainability and readability.
Implementing the WAITAOF functionality which would allow the user to
block until a specified number of Redises have fsynced all previous write
commands to the AOF.
Syntax: `WAITAOF <num_local> <num_replicas> <timeout>`
Response: Array containing two elements: num_local, num_replicas
num_local is always either 0 or 1 representing the local AOF on the master.
num_replicas is the number of replicas that acknowledged the a replication
offset of the last write being fsynced to the AOF.
Returns an error when called on replicas, or when called with non-zero
num_local on a master with AOF disabled, in all other cases the response
just contains number of fsync copies.
Main changes:
* Added code to keep track of replication offsets that are confirmed to have
been fsynced to disk.
* Keep advancing master_repl_offset even when replication is disabled (and
there's no replication backlog, only if there's an AOF enabled).
This way we can use this command and it's mechanisms even when replication
is disabled.
* Extend REPLCONF ACK to `REPLCONF ACK <ofs> FACK <ofs>`, the FACK
will be appended only if there's an AOF on the replica, and already ignored on
old masters (thus backwards compatible)
* WAIT now no longer wait for the replication offset after your last command, but
rather the replication offset after your last write (or read command that caused
propagation, e.g. lazy expiry).
Unrelated changes:
* WAIT command respects CLIENT_DENY_BLOCKING (not just CLIENT_MULTI)
Implementation details:
* Add an atomic var named `fsynced_reploff_pending` that's updated
(usually by the bio thread) and later copied to the main `fsynced_reploff`
variable (only if the AOF base file exists).
I.e. during the initial AOF rewrite it will not be used as the fsynced offset
since the AOF base is still missing.
* Replace close+fsync bio job with new BIO_CLOSE_AOF (AOF specific)
job that will also update fsync offset the field.
* Handle all AOF jobs (BIO_CLOSE_AOF, BIO_AOF_FSYNC) in the same bio
worker thread, to impose ordering on their execution. This solves a
race condition where a job could set `fsynced_reploff_pending` to a higher
value than another pending fsync job, resulting in indicating an offset
for which parts of the data have not yet actually been fsynced.
Imposing an ordering on the jobs guarantees that fsync jobs are executed
in increasing order of replication offset.
* Drain bio jobs when switching `appendfsync` to "always"
This should prevent a write race between updates to `fsynced_reploff_pending`
in the main thread (`flushAppendOnlyFile` when set to ALWAYS fsync), and
those done in the bio thread.
* Drain the pending fsync when starting over a new AOF to avoid race conditions
with the previous AOF offsets overriding the new one (e.g. after switching to
replicate from a new master).
* Make sure to update the fsynced offset at the end of the initial AOF rewrite.
a must in case there are no additional writes that trigger a periodic fsync,
specifically for a replica that does a full sync.
Limitations:
It is possible to write a module and a Lua script that propagate to the AOF and doesn't
propagate to the replication stream. see REDISMODULE_ARGV_NO_REPLICAS and luaRedisSetReplCommand.
These features are incompatible with the WAITAOF command, and can result
in two bad cases. The scenario is that the user executes command that only
propagates to AOF, and then immediately
issues a WAITAOF, and there's no further writes on the replication stream after that.
1. if the the last thing that happened on the replication stream is a PING
(which increased the replication offset but won't trigger an fsync on the replica),
then the client would hang forever (will wait for an fack that the replica will never
send sine it doesn't trigger any fsyncs).
2. if the last thing that happened is a write command that got propagated properly,
then WAITAOF will be released immediately, without waiting for an fsync (since
the offset didn't change)
Refactoring:
* Plumbing to allow bio worker to handle multiple job types
This introduces infrastructure necessary to allow BIO workers to
not have a 1-1 mapping of worker to job-type. This allows in the
future to assign multiple job types to a single worker, either as
a performance/resource optimization, or as a way of enforcing
ordering between specific classes of jobs.
Co-authored-by: Oran Agra <oran@redislabs.com>
This bug seems to be there forever, CLIENT REPLY OFF|SKIP will
mark the client with CLIENT_REPLY_OFF or CLIENT_REPLY_SKIP flags.
With these flags, prepareClientToWrite called by addReply* will
return C_ERR directly. So the client can't receive the Pub/Sub
messages and any other push notifications, e.g client side tracking.
In this PR, we adding a CLIENT_PUSHING flag, disables the reply
silencing flags. When adding push replies, set the flag, after the reply,
clear the flag. Then add the flag check in prepareClientToWrite.
Fixes#11874
Note, the SUBSCRIBE command response is a bit awkward,
see https://github.com/redis/redis-doc/pull/2327
Co-authored-by: Oran Agra <oran@redislabs.com>
Work in progress towards implementing a reply schema as part of COMMAND DOCS, see #9845
Since ironing the details of the reply schema of each and every command can take a long time, we
would like to merge this PR when the infrastructure is ready, and let this mature in the unstable branch.
Meanwhile the changes of this PR are internal, they are part of the repo, but do not affect the produced build.
### Background
In #9656 we add a lot of information about Redis commands, but we are missing information about the replies
### Motivation
1. Documentation. This is the primary goal.
2. It should be possible, based on the output of COMMAND, to be able to generate client code in typed
languages. In order to do that, we need Redis to tell us, in detail, what each reply looks like.
3. We would like to build a fuzzer that verifies the reply structure (for now we use the existing
testsuite, see the "Testing" section)
### Schema
The idea is to supply some sort of schema for the various replies of each command.
The schema will describe the conceptual structure of the reply (for generated clients), as defined in RESP3.
Note that the reply structure itself may change, depending on the arguments (e.g. `XINFO STREAM`, with
and without the `FULL` modifier)
We decided to use the standard json-schema (see https://json-schema.org/) as the reply-schema.
Example for `BZPOPMIN`:
```
"reply_schema": {
"oneOf": [
{
"description": "Timeout reached and no elements were popped.",
"type": "null"
},
{
"description": "The keyname, popped member, and its score.",
"type": "array",
"minItems": 3,
"maxItems": 3,
"items": [
{
"description": "Keyname",
"type": "string"
},
{
"description": "Member",
"type": "string"
},
{
"description": "Score",
"type": "number"
}
]
}
]
}
```
#### Notes
1. It is ok that some commands' reply structure depends on the arguments and it's the caller's responsibility
to know which is the relevant one. this comes after looking at other request-reply systems like OpenAPI,
where the reply schema can also be oneOf and the caller is responsible to know which schema is the relevant one.
2. The reply schemas will describe RESP3 replies only. even though RESP3 is structured, we want to use reply
schema for documentation (and possibly to create a fuzzer that validates the replies)
3. For documentation, the description field will include an explanation of the scenario in which the reply is sent,
including any relation to arguments. for example, for `ZRANGE`'s two schemas we will need to state that one
is with `WITHSCORES` and the other is without.
4. For documentation, there will be another optional field "notes" in which we will add a short description of
the representation in RESP2, in case it's not trivial (RESP3's `ZRANGE`'s nested array vs. RESP2's flat
array, for example)
Given the above:
1. We can generate the "return" section of all commands in [redis-doc](https://redis.io/commands/)
(given that "description" and "notes" are comprehensive enough)
2. We can generate a client in a strongly typed language (but the return type could be a conceptual
`union` and the caller needs to know which schema is relevant). see the section below for RESP2 support.
3. We can create a fuzzer for RESP3.
### Limitations (because we are using the standard json-schema)
The problem is that Redis' replies are more diverse than what the json format allows. This means that,
when we convert the reply to a json (in order to validate the schema against it), we lose information (see
the "Testing" section below).
The other option would have been to extend the standard json-schema (and json format) to include stuff
like sets, bulk-strings, error-string, etc. but that would mean also extending the schema-validator - and that
seemed like too much work, so we decided to compromise.
Examples:
1. We cannot tell the difference between an "array" and a "set"
2. We cannot tell the difference between simple-string and bulk-string
3. we cannot verify true uniqueness of items in commands like ZRANGE: json-schema doesn't cover the
case of two identical members with different scores (e.g. `[["m1",6],["m1",7]]`) because `uniqueItems`
compares (member,score) tuples and not just the member name.
### Testing
This commit includes some changes inside Redis in order to verify the schemas (existing and future ones)
are indeed correct (i.e. describe the actual response of Redis).
To do that, we added a debugging feature to Redis that causes it to produce a log of all the commands
it executed and their replies.
For that, Redis needs to be compiled with `-DLOG_REQ_RES` and run with
`--reg-res-logfile <file> --client-default-resp 3` (the testsuite already does that if you run it with
`--log-req-res --force-resp3`)
You should run the testsuite with the above args (and `--dont-clean`) in order to make Redis generate
`.reqres` files (same dir as the `stdout` files) which contain request-response pairs.
These files are later on processed by `./utils/req-res-log-validator.py` which does:
1. Goes over req-res files, generated by redis-servers, spawned by the testsuite (see logreqres.c)
2. For each request-response pair, it validates the response against the request's reply_schema
(obtained from the extended COMMAND DOCS)
5. In order to get good coverage of the Redis commands, and all their different replies, we chose to use
the existing redis test suite, rather than attempt to write a fuzzer.
#### Notes about RESP2
1. We will not be able to use the testing tool to verify RESP2 replies (we are ok with that, it's time to
accept RESP3 as the future RESP)
2. Since the majority of the test suite is using RESP2, and we want the server to reply with RESP3
so that we can validate it, we will need to know how to convert the actual reply to the one expected.
- number and boolean are always strings in RESP2 so the conversion is easy
- objects (maps) are always a flat array in RESP2
- others (nested array in RESP3's `ZRANGE` and others) will need some special per-command
handling (so the client will not be totally auto-generated)
Example for ZRANGE:
```
"reply_schema": {
"anyOf": [
{
"description": "A list of member elements",
"type": "array",
"uniqueItems": true,
"items": {
"type": "string"
}
},
{
"description": "Members and their scores. Returned in case `WITHSCORES` was used.",
"notes": "In RESP2 this is returned as a flat array",
"type": "array",
"uniqueItems": true,
"items": {
"type": "array",
"minItems": 2,
"maxItems": 2,
"items": [
{
"description": "Member",
"type": "string"
},
{
"description": "Score",
"type": "number"
}
]
}
}
]
}
```
### Other changes
1. Some tests that behave differently depending on the RESP are now being tested for both RESP,
regardless of the special log-req-res mode ("Pub/Sub PING" for example)
2. Update the history field of CLIENT LIST
3. Added basic tests for commands that were not covered at all by the testsuite
### TODO
- [x] (maybe a different PR) add a "condition" field to anyOf/oneOf schemas that refers to args. e.g.
when `SET` return NULL, the condition is `arguments.get||arguments.condition`, for `OK` the condition
is `!arguments.get`, and for `string` the condition is `arguments.get` - https://github.com/redis/redis/issues/11896
- [x] (maybe a different PR) also run `runtest-cluster` in the req-res logging mode
- [x] add the new tests to GH actions (i.e. compile with `-DLOG_REQ_RES`, run the tests, and run the validator)
- [x] (maybe a different PR) figure out a way to warn about (sub)schemas that are uncovered by the output
of the tests - https://github.com/redis/redis/issues/11897
- [x] (probably a separate PR) add all missing schemas
- [x] check why "SDOWN is triggered by misconfigured instance replying with errors" fails with --log-req-res
- [x] move the response transformers to their own file (run both regular, cluster, and sentinel tests - need to
fight with the tcl including mechanism a bit)
- [x] issue: module API - https://github.com/redis/redis/issues/11898
- [x] (probably a separate PR): improve schemas: add `required` to `object`s - https://github.com/redis/redis/issues/11899
Co-authored-by: Ozan Tezcan <ozantezcan@gmail.com>
Co-authored-by: Hanna Fadida <hanna.fadida@redislabs.com>
Co-authored-by: Oran Agra <oran@redislabs.com>
Co-authored-by: Shaya Potter <shaya@redislabs.com>
This PR has two parts:
1. Fix flaky test case, the previous tests set a lot of volatile keys,
it injects an unexpected DEL command into the replication stream during
the later test, causing it to fail. Add a flushall to avoid it.
2. Improve assert_replication_stream, now it can print the whole stream
rather than just the failing line.
Currently (starting at #11012) When a module is blocked on keys it sets the
CLIENT_PENDING_COMMAND flag.
However in case the module decides to unblock the client not via the regular flow
(eg timeout, key signal or CLIENT UNBLOCK command) it will attempt to reprocess the
module command and potentially blocked again.
This fix remove the CLIENT_PENDING_COMMAND flag in case blockedForKeys is
issued from module context.
I've seen it fail here (test-centos7-tls-module-no-tls and test-freebsd):
```
*** [err]: Operations in no-touch mode do not alter the last access time of a key in tests/unit/introspection-2.tcl
Expected '244296' to be more than '244296' (context: type eval line 12 cmd {assert_morethan $newlru $oldlru} proc ::test)
```
Our LRU_CLOCK_RESOLUTION value is 1000ms, and default hz is 10, so if the
test is really fast, or the timing is just right, newlru will be the same
as oldlru. We fixed this by changing `after 1000` to `after 1100`.
Test `trim on SET with big value` (introduced from #11817) fails under mac m1 with libc mem_allocator.
The reason is that malloc(33000) will allocate 65536 bytes(>42000).
This test still passes under ubuntu with libc mem_allocator.
```
*** [err]: trim on SET with big value in tests/unit/type/string.tcl
Expected [r memory usage key] < 42000 (context: type source line 471 file /Users/iospack/data/redis_fork/tests/unit/type/string.tcl cmd {assert {[r memory usage key] < 42000}} proc ::test)
```
simple test under mac m1 with libc mem_allocator:
```c
void *p = zmalloc(33000);
printf("malloc size: %zu\n", zmalloc_size(p));
# output
malloc size: 65536
```
This test is very sensitive and fragile. It often fails in Daily,
in most cases, it failed in test-ubuntu-32bit (the AOF loading one),
with the range in (31, 40):
```
[err]: Active defrag in tests/unit/memefficiency.tcl
Expected 38 <= 30 (context: type eval line 113 cmd {assert {$max_latency <= 30}} proc ::test)
```
The AOF loading part isn't tightly fixed to the cron hz. It calls
processEventsWhileBlocked once in every 1024 command calls.
```
/* Serve the clients from time to time */
if (!(loops++ % 1024)) {
off_t progress_delta = ftello(fp) - last_progress_report_size;
loadingIncrProgress(progress_delta);
last_progress_report_size += progress_delta;
processEventsWhileBlocked();
processModuleLoadingProgressEvent(1);
}
```
In this case, we can either decrease the 1024 or increase the
threshold of just the AOF part of that test. Considering the test
machines are sometimes slow, and all sort of quirks could happen
(which do not indicate a bug), and we've already set to 30, we suppose
we can set it a little bit higher, set it to 40. We can have this instead of
adding another testing config (we can add it when we really need it).
Fixes#11868
As `sdsRemoveFreeSpace` have an impact on performance even if it is a no-op (see details at #11508).
Only call the function when there is a possibility that the string contains free space.
* For strings coming from the network, it's only if they're bigger than PROTO_MBULK_BIG_ARG
* For strings coming from scripts, it's only if they're smaller than LUA_CMD_OBJCACHE_MAX_LEN
* For strings coming from modules, it could be anything.
Co-authored-by: Oran Agra <oran@redislabs.com>
Co-authored-by: sundb <sundbcn@gmail.com>
Authenticated users can use string matching commands with a
specially crafted pattern to trigger a denial-of-service attack on Redis,
causing it to hang and consume 100% CPU time.
Co-authored-by: Tom Levy <tomlevy93@gmail.com>
Avoid calling unwatchAllKeys when running touchAllWatchedKeysInDb (which was unnecessary)
This can potentially lead to use-after-free and memory corruption when the next entry
pointer held by the watched keys iterator is freed when unwatching all keys of a specific client.
found with address sanitizer, added a test which will not always fail (depending on the random
dict hashing seed)
problem introduced in #9829 (Reids 7.0)
Co-authored-by: Oran Agra <oran@redislabs.com>
The test sporadically failed with valgrind trying to match
`no client named obuf-client1 found*`
in the log it looks like `obuf-client1` was indeed dropped,
so i'm guessing it's because CLIENT LIST was processed first.
When no-touch mode is enabled, the client will not touch LRU/LFU of the
keys it accesses, except when executing command `TOUCH`.
This allows inspecting or modifying the key-space without affecting their eviction.
Changes:
- A command `CLIENT NO-TOUCH ON|OFF` to switch on and off this mode.
- A client flag `#define CLIENT_NOTOUCH (1ULL<<45)`, which can be shown
with `CLIENT INFO`, by the letter "T" in the "flags" field.
- Clear `NO-TOUCH` flag in `clearClientConnectionState`, which is used by `RESET`
command and resetting temp clients used by modules.
- Also clear `NO-EVICT` flag in `clearClientConnectionState`, this might have been an
oversight, spotted by @madolson.
- A test using `DEBUG OBJECT` command to verify that LRU stat is not touched when
no-touch mode is on.
Co-authored-by: chentianjie <chentianjie@alibaba-inc.com>
Co-authored-by: Madelyn Olson <34459052+madolson@users.noreply.github.com>
Co-authored-by: sundb <sundbcn@gmail.com>
We noticed that `client evicted due to client tracking prefixes`
takes over 200 seconds with valgrind.
We combine three prefixes in each command, this will probably
save us half the testing time.
Before: normal: 3508ms, valgrind: 289503ms -> 290s
With three prefixes, normal: 1500ms, valgrind: 135742ms -> 136s
Since we did not actually count the memory usage of all prefixes, see
getClientMemoryUsage, so we can not use larger prefixes to speed up the
test here. Also this PR cleaned up some spaces (IDE jobs) and typos.
There is a built in limit to client side tracking keys, which when exceeded will invalidate keys. This occurs in two places, one in the server cron and other before executing a command. If it happens in the second scenario, the invalidations will be queued for later since current client is set. This queue is never drained if a command is not executed (through call) such as a multi-exec command getting queued. This results in a later server assert crashing.
This test case is to cover a edge scenario: when a writable replica enabled AOF
at the same time, active expiry keys which was created in writable replicas should
propagate to the AOF file, and some versions might crash (fixed by #11615).
For details, please refer to #11778
* Make it clear that current_client is the root client that was called by
external connection
* add executing_client which is the client that runs the current command
(can be a module or a script)
* Remove script_caller that was used for commands that have CLIENT_SCRIPT
to get the client that called the script. in most cases, that's the current_client,
and in others (when being called from a module), it could be an intermediate
client when we actually want the original one used by the external connection.
bugfixes:
* RM_Call with C flag should log ACL errors with the requested user rather than
the one used by the original client, this also solves a crash when RM_Call is used
with C flag from a detached thread safe context.
* addACLLogEntry would have logged info about the script_caller, but in case the
script was issued by a module command we actually want the current_client. the
exception is when RM_Call is called from a timer event, in which case we don't
have a current_client.
behavior changes:
* client side tracking for scripts now tracks the keys that are read by the script
instead of the keys that are declared by the caller for EVAL
other changes:
* Log both current_client and executing_client in the crash log.
* remove prepareLuaClient and resetLuaClient, being dead code that was forgotten.
* remove scriptTimeSnapshot and snapshot_time and instead add cmd_time_snapshot
that serves all commands and is reset only when execution nesting starts.
* remove code to propagate CLIENT_FORCE_REPL from the executed command
to the script caller since scripts aren't propagated anyway these days and anyway
this flag wouldn't have had an effect since CLIENT_PREVENT_PROP is added by scriptResetRun.
* fix a module GIL violation issue in afterSleep that was introduced in #10300 (unreleased)
In #9373, actually need to replace `$rd $pop blist1{t} blist2{t} 1`
with `bpop_command_two_key $rd $pop blist1{t} blist2{t} 1` but forgot
to delete the latter.
This doesn't affect the test, because the later assert_error "WRONGTYPE"
is expected (and right). And if we read $rd again, it will get the
wrong result, like 'ERR unknown command 'BLMPOP_LEFT' | 'BLMPOP_RIGHT'
All of the POP commands must not decr length below 0.
So, get_fsl will delete the key if the length is 0 (unless
the caller wished to create if doesn't exist)
Other:
1. Use REDISMODULE_WRITE where needed (POP commands)
2. Use wait_for_blokced_clients in test
Unrelated:
Use quotes instead of curly braces in zset.tcl, for variable expansion
Starting from Redis 7.0 (#9890) we started wrapping everything a command
propagates with MULTI/EXEC. The problem is that both SCAN and RANDOMKEY can
lazy-expire arbitrary keys (similar behavior to active-expire), and put DELs in a transaction.
Fix: When these commands are called without a parent exec-unit (e.g. not in EVAL or
MULTI) we avoid wrapping their DELs in a transaction (for the same reasons active-expire
and eviction avoids a transaction)
This PR adds a per-command flag that indicates that the command may touch arbitrary
keys (not the ones in the arguments), and uses that flag to avoid the MULTI-EXEC.
For now, this flag is internal, since we're considering other solutions for the future.
Note for cluster mode: if SCAN/RANDOMKEY is inside EVAL/MULTI it can still cause the
same situation (as it always did), but it won't cause a CROSSSLOT because replicas and AOF
do not perform slot checks.
The problem with the above is mainly for 3rd party ecosystem tools that propagate commands
from master to master, or feed an AOF file with redis-cli into a master.
This PR aims to fix the regression in redis 7.0, and we opened #11792 to try to handle the
bigger problem with lazy expire better for another release.
The PR adds support for the following flags on RedisModule_OpenKey:
* REDISMODULE_OPEN_KEY_NONOTIFY - Don't trigger keyspace event on key misses.
* REDISMODULE_OPEN_KEY_NOSTATS - Don't update keyspace hits/misses counters.
* REDISMODULE_OPEN_KEY_NOEXPIRE - Avoid deleting lazy expired keys.
* REDISMODULE_OPEN_KEY_NOEFFECTS - Avoid any effects from fetching the key
In addition, added `RM_GetOpenKeyModesAll`, which returns the mask of all
supported OpenKey modes. This allows the module to check, in runtime, which
OpenKey modes are supported by the current Redis instance.
Return an error when loadAppendOnlyFiles fails instead of
exiting. DEBUF LOADAOF command is only meant to be used by
the test suite, and only by tests that generated an AOF file
first. So this change is ok (considering that the caller is
likely to catch this error and die).
This actually revert part of the code in #9012, and now
DEBUG LOADAOF behaves the same as DEBUG RELOAD (returns an
error when the load fails).
Plus remove a `after 2000` in a test, which can save times (looks like copy paste error).
There is a timing issue in the new ACL log test:
```
*** [err]: ACL LOG aggregates similar errors together and assigns unique entry-id to new errors in tests/unit/acl.tcl
Expected 1675382873989 < 1675382873989 (context: type eval line 15 cmd {assert {$timestamp_last_update_original < $timestamp_last_updated_after_update}} proc ::test)
```
Looking at the test code, we will check the `timestamp-last-updated` before
and after a new ACL error occurs. Actually `WRONGPASS` errors can be executed
very quickly on fast machines. For example, in the this case, the execution is
completed within one millisecond.
The error is easy to reproduce, if we reduce the number of the for loops, for
example set to 2, and using --loop and --stop. Avoid this timing issue by adding
an `after 1` before the new errors.
The test was introduced in #11477.
Added 3 fields to the ACL LOG - adds entry_id, timestamp_created and timestamp_last_updated, which updates similar existing log error entries. The pair - entry_id, timestamp_created is a unique identifier of this entry, in case the node dies and is restarted, it can detect that if it's a new series.
The primary use case of Unique id is to uniquely identify the error messages and not to detect if the server has restarted.
entry-id is the sequence number of the entry (starting at 0) since the server process started. Can also be used to check if items were "lost" if they fell between periods.
timestamp-created is the unix-time in ms at the time the entry was first created.
timestamp-last-updated is the unix-time in ms at the time the entry was last updated
Time_created gives the absolute time which better accounts for network time as compared to time since. It can also be older than 60 secs and presently there is no field that can display the original time of creation once the error entry is updated.
The reason of timestamp_last_updated field is that it provides a more precise value for the “last time” an error was seen where as, presently it is only in the 60 second period.
Co-authored-by: Madelyn Olson <madelyneolson@gmail.com>
Currently while a sharded pubsub message publish tries to propagate the message across the cluster, a NULL check is missing for clusterLink. clusterLink could be NULL if the link is causing memory beyond the set threshold cluster-link-sendbuf-limit and server terminates the link.
This change introduces two things:
Avoids the engine crashes on the publishing node if a message is tried to be sent to a node and the link is NULL.
Adds a debugging tool CLUSTERLINK KILL to terminate the clusterLink between two nodes.
In #7875 (Redis 6.2), we changed the sds alloc to be the usable allocation
size in order to:
> reduce the need for realloc calls by making the sds implicitly take over
the internal fragmentation
This change was done most sds functions, excluding `sdsRemoveFreeSpace` and
`sdsResize`, the reason is that in some places (e.g. clientsCronResizeQueryBuffer)
we call sdsRemoveFreeSpace when we see excessive free space and want to trim it.
so if we don't trim it exactly to size, the caller may still see excessive free space and
call it again and again.
However, this resulted in some excessive calls to realloc, even when there's no need
and it's gonna be a no-op (e.g. when reducing 15 bytes allocation to 13).
It turns out that a call for realloc with jemalloc can be expensive even if it ends up
doing nothing, so this PR adds a check using `je_nallocx`, which is cheap to avoid
the call for realloc.
in addition to that this PR unifies sdsResize and sdsRemoveFreeSpace into common
code. the difference between them was that sdsResize would avoid using SDS_TYPE_5,
since it want to keep the string ready to be resized again, while sdsRemoveFreeSpace
would permit using SDS_TYPE_5 and get an optimal memory consumption.
now both methods take a `would_regrow` argument that makes it more explicit.
the only actual impact of that is that in clientsCronResizeQueryBuffer we call both sdsResize
and sdsRemoveFreeSpace for in different cases, and we now prevent the use of SDS_TYPE_5 in both.
The new test that was added to cover this concern used to pass before this PR as well,
this PR is just a performance optimization and cleanup.
Benchmark:
`redis-benchmark -c 100 -t set -d 512 -P 10 -n 100000000`
on i7-9850H with jemalloc, shows improvement from 1021k ops/sec to 1067k (average of 3 runs).
some 4.5% improvement.
Co-authored-by: Oran Agra <oran@redislabs.com>
