# Description
Gather most of the scattered `redisDb`-related code from the per-slot
dict PR (#11695) and turn it to a new data structure, `kvstore`. i.e.
it's a class that represents an array of dictionaries.
# Motivation
The main motivation is code cleanliness, the idea of using an array of
dictionaries is very well-suited to becoming a self-contained data
structure.
This allowed cleaning some ugly code, among others: loops that run twice
on the main dict and expires dict, and duplicate code for allocating and
releasing this data structure.
# Notes
1. This PR reverts the part of https://github.com/redis/redis/pull/12848
where the `rehashing` list is global (handling rehashing `dict`s is
under the responsibility of `kvstore`, and should not be managed by the
server)
2. This PR also replaces the type of `server.pubsubshard_channels` from
`dict**` to `kvstore` (original PR:
https://github.com/redis/redis/pull/12804). After that was done,
server.pubsub_channels was also chosen to be a `kvstore` (with only one
`dict`, which seems odd) just to make the code cleaner by making it the
same type as `server.pubsubshard_channels`, see
`pubsubtype.serverPubSubChannels`
3. the keys and expires kvstores are currenlty configured to allocate
the individual dicts only when the first key is added (unlike before, in
which they allocated them in advance), but they won't release them when
the last key is deleted.
Worth mentioning that due to the recent change the reply of DEBUG
HTSTATS changed, in case no keys were ever added to the db.
before:
```
127.0.0.1:6379> DEBUG htstats 9
[Dictionary HT]
Hash table 0 stats (main hash table):
No stats available for empty dictionaries
[Expires HT]
Hash table 0 stats (main hash table):
No stats available for empty dictionaries
```
after:
```
127.0.0.1:6379> DEBUG htstats 9
[Dictionary HT]
[Expires HT]
```
When db->expires_cursor==0, it means the DB is done the scanning,
we should exit the loop to avoid the useless scanning.
It is easy to see the active expire timeout in the modified test,
for example, let's assume that there is only 1 expired key in the
DB, and the size / buckets ratio is less than 1%, which means that
we will skip it in isExpiryDictValidForSamplingCb, and the return
value of expires_cursor is 0.
Because `data.sampled == 0` is always true, so `repeat` is also
always true, we will keep scanning the DB, but every time it is
skipped by the previous judgment (expires_cursor = 0), until the
timelimit is finally exhausted.
Currently, We compute `db->avg_ttl` after each short `dbScan` sweep (a
few buckets without checking the time limit). But after each `dbScan`
sweep, we don't have much data and this makes the db->avg_ttl less
precise. For example, even if we scan the whole db, we can't get the
exact avg_ttl because we separate the data.
i.e. because of the running average, if we issue 16 calls to scan, we'll
give lower weight to the first one, and higher weight to the last one.
I think we should calculate `db->avg_ttl` until completing more of the
db iteration (judgement of time limit or the beginning of iterating next
db) because we have more sample data in this db and can get more
accurate result. In the best case, if we scan the whole db, we can get
the exact avg_ttl.
In this PR, we postpone the avg_ttl calculation until the judgement of
time limit or iteration of next db, so we can accumulate more data to
get more precise avg_ttl.
Note that we still need to make sure to decay the old TTLs at the same
speed as before, which is why we want to run the decay mechanism several
times, or use the Pow formula, see the comment in the code.
In my experiment, this PR can improve 89% or 52% accuracy in different
workload.
Co-authored-by: Oran Agra <oran@redislabs.com>
Currently, when the number of databases exceeds 16,
the efficiency of cleaning expired keys is relatively low.
The reason is that by default only 16 databases are scanned when
attempting to clean expired keys (CRON_DBS_PER_CALL is 16). But users
may set databases higher than 16, such as 256, but it does not
necessarily mean that all 256 databases have expiration time set. If
only one database has expiration time set, this database needs 16
activeExpireCycle rounds in order to be scanned once, and 15 of those
rounds are meaningless.
To optimize the efficiency of expiration in such scenarios, we use dbs_per_call
to control the number of databases with expired keys being scanned.
Additionally, add a condition to limit the maximum number of rounds
to server.dbnum to prevent excessive spinning. This ensures that even if
only one database has expired keys, it can be triggered within a single cron job.
Redis 7.2 (#9406) introduced a new modules event, `RedisModuleEvent_Key`.
This new event allows the module to read the key data just before it is removed
from the database (either deleted, expired, evicted, or overwritten).
When the key is removed from the database, either by active expire or eviction.
The new event was not called as part of an execution unit. This can cause an
issue if the module registers a post notification job inside the event. This job will
not be executed atomically with the expiration/eviction operation and will not
replicated inside a Multi/Exec. Moreover, the post notification job will be executed
right after the event where it is still not safe to perform any write operation, this will
violate the promise that post notification job will be called atomically with the
operation that triggered it and **only when it is safe to write**.
This PR fixes the issue by wrapping each expiration/eviction of a key with an execution
unit. This makes sure the entire operation will run atomically and all the post notification
jobs will be executed at the end where it is safe to write.
Tests were modified to verify the fix.
Optimize the performance of SCAN commands when a match pattern can only contain keys from a
single slot in cluster mode. This can happen when the pattern contains a hash tag before any
wildcard matchers or when the key contains no matchers.
This is an implementation of https://github.com/redis/redis/issues/10589 that eliminates 16 bytes per entry in cluster mode, that are currently used to create a linked list between entries in the same slot. Main idea is splitting main dictionary into 16k smaller dictionaries (one per slot), so we can perform all slot specific operations, such as iteration, without any additional info in the `dictEntry`. For Redis cluster, the expectation is that there will be a larger number of keys, so the fixed overhead of 16k dictionaries will be The expire dictionary is also split up so that each slot is logically decoupled, so that in subsequent revisions we will be able to atomically flush a slot of data.
## Important changes
* Incremental rehashing - one big change here is that it's not one, but rather up to 16k dictionaries that can be rehashing at the same time, in order to keep track of them, we introduce a separate queue for dictionaries that are rehashing. Also instead of rehashing a single dictionary, cron job will now try to rehash as many as it can in 1ms.
* getRandomKey - now needs to not only select a random key, from the random bucket, but also needs to select a random dictionary. Fairness is a major concern here, as it's possible that keys can be unevenly distributed across the slots. In order to address this search we introduced binary index tree). With that data structure we are able to efficiently find a random slot using binary search in O(log^2(slot count)) time.
* Iteration efficiency - when iterating dictionary with a lot of empty slots, we want to skip them efficiently. We can do this using same binary index that is used for random key selection, this index allows us to find a slot for a specific key index. For example if there are 10 keys in the slot 0, then we can quickly find a slot that contains 11th key using binary search on top of the binary index tree.
* scan API - in order to perform a scan across the entire DB, the cursor now needs to not only save position within the dictionary but also the slot id. In this change we append slot id into LSB of the cursor so it can be passed around between client and the server. This has interesting side effect, now you'll be able to start scanning specific slot by simply providing slot id as a cursor value. The plan is to not document this as defined behavior, however. It's also worth nothing the SCAN API is now technically incompatible with previous versions, although practically we don't believe it's an issue.
* Checksum calculation optimizations - During command execution, we know that all of the keys are from the same slot (outside of a few notable exceptions such as cross slot scripts and modules). We don't want to compute the checksum multiple multiple times, hence we are relying on cached slot id in the client during the command executions. All operations that access random keys, either should pass in the known slot or recompute the slot.
* Slot info in RDB - in order to resize individual dictionaries correctly, while loading RDB, it's not enough to know total number of keys (of course we could approximate number of keys per slot, but it won't be precise). To address this issue, we've added additional metadata into RDB that contains number of keys in each slot, which can be used as a hint during loading.
* DB size - besides `DBSIZE` API, we need to know size of the DB in many places want, in order to avoid scanning all dictionaries and summing up their sizes in a loop, we've introduced a new field into `redisDb` that keeps track of `key_count`. This way we can keep DBSIZE operation O(1). This is also kept for O(1) expires computation as well.
## Performance
This change improves SET performance in cluster mode by ~5%, most of the gains come from us not having to maintain linked lists for keys in slot, non-cluster mode has same performance. For workloads that rely on evictions, the performance is similar because of the extra overhead for finding keys to evict.
RDB loading performance is slightly reduced, as the slot of each key needs to be computed during the load.
## Interface changes
* Removed `overhead.hashtable.slot-to-keys` to `MEMORY STATS`
* Scan API will now require 64 bits to store the cursor, even on 32 bit systems, as the slot information will be stored.
* New RDB version to support the new op code for SLOT information.
---------
Co-authored-by: Vitaly Arbuzov <arvit@amazon.com>
Co-authored-by: Harkrishn Patro <harkrisp@amazon.com>
Co-authored-by: Roshan Khatri <rvkhatri@amazon.com>
Co-authored-by: Madelyn Olson <madelyneolson@gmail.com>
Co-authored-by: Oran Agra <oran@redislabs.com>
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
This change deletes the dictGetNext and dictGetNextRef functions, so the
dict API doesn't expose the next field at all.
The bucket function in dictScan is deleted. A separate dictScanDefrag function
is added which takes a defrag alloc function to defrag-reallocate the dict entries.
"Dirty" code accessing the dict internals in active defrag is removed.
An 'afterReplaceEntry' is added to dictType, which allows the dict user
to keep the dictEntry metadata up to date after reallocation/defrag/move.
Additionally, for updating the cluster slot-to-key mapping, after a dictEntry
has been reallocated, we need to know which db a dict belongs to, so we store
a pointer to the db in a new metadata section in the dict struct, which is
a new mechanism similar to dictEntry metadata. This adds some complexity but
provides better isolation.
1. Get rid of server.core_propagates - we can just rely on module/call nesting levels
2. Rename in_nested_call to execution_nesting and update the comment
3. Remove module_ctx_nesting (redundant, we can use execution_nesting)
4. Modify postExecutionUnitOperations according to the comment (The main purpose of this PR)
5. trackingHandlePendingKeyInvalidations: Check the nesting level inside this function
We need to honor the post-execution-unit API and call it after each KSN
Note that this is an edge case that only happens in case volatile keys were
created directly on a writable replica, and that anyway nothing is propagated to sub-replicas
Co-authored-by: Oran Agra <oran@redislabs.com>
There is overhead on Redis 7.0 EXPIRE command that is not present on 6.2.7.
We could see that on the unstable profile there are around 7% of CPU cycles
spent on rewriteClientCommandVector that are not present on 6.2.7.
This was introduced in #8474.
This PR reduces the overhead by using 2X rewriteClientCommandArgument instead of
rewriteClientCommandVector. In this scenario rewriteClientCommandVector creates 4 arguments.
the above usage of rewriteClientCommandArgument reduces the overhead in half.
This PR should also improve PEXPIREAT performance by avoiding at all
rewriteClientCommandArgument usage.
Co-authored-by: Oran Agra <oran@redislabs.com>
Add a new module event `RedisModule_Event_Key`, this event is fired
when a key is removed from the keyspace.
The event includes an open key that can be used for reading the key before
it is removed. Modules can also extract the key-name, and use RM_Open
or RM_Call to access key from within that event, but shouldn't modify anything
from within this event.
The following sub events are available:
- `REDISMODULE_SUBEVENT_KEY_DELETED`
- `REDISMODULE_SUBEVENT_KEY_EXPIRED`
- `REDISMODULE_SUBEVENT_KEY_EVICTED`
- `REDISMODULE_SUBEVENT_KEY_OVERWRITE`
The data pointer can be casted to a RedisModuleKeyInfo structure
with the following fields:
```
RedisModuleKey *key; // Opened Key
```
### internals
* We also add two dict functions:
`dictTwoPhaseUnlinkFind` finds an element from the table, also get the plink of the entry.
The entry is returned if the element is found. The user should later call `dictTwoPhaseUnlinkFree`
with it in order to unlink and release it. Otherwise if the key is not found, NULL is returned.
These two functions should be used in pair. `dictTwoPhaseUnlinkFind` pauses rehash and
`dictTwoPhaseUnlinkFree` resumes rehash.
* We change `dbOverwrite` to `dbReplaceValue` which just replaces the value of the key and
doesn't fire any events. The "overwrite" part (which emits events) is just when called from `setKey`,
the other places that called dbOverwrite were ones that just update the value in-place (INCR*, SPOP,
and dbUnshareStringValue). This should not have any real impact since `moduleNotifyKeyUnlink` and
`signalDeletedKeyAsReady` wouldn't have mattered in these cases anyway (i.e. module keys and
stream keys didn't have direct calls to dbOverwrite)
* since we allow doing RM_OpenKey from withing these callbacks, we temporarily disable lazy expiry.
* We also temporarily disable lazy expiry when we are in unlink/unlink2 callback and keyspace
notification callback.
* Move special definitions to the top of redismodule.h
This is needed to resolve compilation errors with RedisModuleKeyInfoV1
that carries a RedisModuleKey member.
Co-authored-by: Oran Agra <oran@redislabs.com>
### Summary of API additions
* `RedisModule_AddPostNotificationJob` - new API to call inside a key space
notification (and on more locations in the future) and allow to add a post job as describe above.
* New module option, `REDISMODULE_OPTIONS_ALLOW_NESTED_KEYSPACE_NOTIFICATIONS`,
allows to disable Redis protection of nested key-space notifications.
* `RedisModule_GetModuleOptionsAll` - gets the mask of all supported module options so a module
will be able to check if a given option is supported by the current running Redis instance.
### Background
The following PR is a proposal of handling write operations inside module key space notifications.
After a lot of discussions we came to a conclusion that module should not perform any write
operations on key space notification.
Some examples of issues that such write operation can cause are describe on the following links:
* Bad replication oreder - https://github.com/redis/redis/pull/10969
* Used after free - https://github.com/redis/redis/pull/10969#issuecomment-1223771006
* Used after free - https://github.com/redis/redis/pull/9406#issuecomment-1221684054
There are probably more issues that are yet to be discovered. The underline problem with writing
inside key space notification is that the notification runs synchronously, this means that the notification
code will be executed in the middle on Redis logic (commands logic, eviction, expire).
Redis **do not assume** that the data might change while running the logic and such changes
can crash Redis or cause unexpected behaviour.
The solution is to state that modules **should not** perform any write command inside key space
notification (we can chose whether or not we want to force it). To still cover the use-case where
module wants to perform a write operation as a reaction to key space notifications, we introduce
a new API , `RedisModule_AddPostNotificationJob`, that allows to register a callback that will be
called by Redis when the following conditions hold:
* It is safe to perform any write operation.
* The job will be called atomically along side the operation that triggers it (in our case, key
space notification).
Module can use this new API to safely perform any write operation and still achieve atomicity
between the notification and the write.
Although currently the API is supported on key space notifications, the API is written in a generic
way so that in the future we will be able to use it on other places (server events for example).
### Technical Details
Whenever a module uses `RedisModule_AddPostNotificationJob` the callback is added to a list
of callbacks (called `modulePostExecUnitJobs`) that need to be invoke after the current execution
unit ends (whether its a command, eviction, or active expire). In order to trigger those callback
atomically with the notification effect, we call those callbacks on `postExecutionUnitOperations`
(which was `propagatePendingCommands` before this PR). The new function fires the post jobs
and then calls `propagatePendingCommands`.
If the callback perform more operations that triggers more key space notifications. Those keys
space notifications might register more callbacks. Those callbacks will be added to the end
of `modulePostExecUnitJobs` list and will be invoke atomically after the current callback ends.
This raises a concerns of entering an infinite loops, we consider infinite loops as a logical bug
that need to be fixed in the module, an attempt to protect against infinite loops by halting the
execution could result in violation of the feature correctness and so **Redis will make no attempt
to protect the module from infinite loops**
In addition, currently key space notifications are not nested. Some modules might want to allow
nesting key-space notifications. To allow that and keep backward compatibility, we introduce a
new module option called `REDISMODULE_OPTIONS_ALLOW_NESTED_KEYSPACE_NOTIFICATIONS`.
Setting this option will disable the Redis key-space notifications nesting protection and will
pass this responsibility to the module.
### Redis infrastructure
This PR promotes the existing `propagatePendingCommands` to an "Execution Unit" concept,
which is called after each atomic unit of execution,
Co-authored-by: Oran Agra <oran@redislabs.com>
Co-authored-by: Yossi Gottlieb <yossigo@gmail.com>
Co-authored-by: Madelyn Olson <34459052+madolson@users.noreply.github.com>
Renamed from "Pause Clients" to "Pause Actions" since the mechanism can pause
several actions in redis, not just clients (e.g. eviction, expiration).
Previously each pause purpose (which has a timeout that's tracked separately from others purposes),
also implicitly dictated what it pauses (reads, writes, eviction, etc). Now it is explicit, and
the actions that are paused (bit flags) are defined separately from the purpose.
- Previously, when using feature pause-client it also implicitly means to make the server static:
- Pause replica traffic
- Pauses eviction processing
- Pauses expire processing
Making the server static is used also for failover and shutdown. This PR internally rebrand
pause-client API to become pause-action API. It also Simplifies pauseClients structure
by replacing pointers array with static array.
The context of this PR is to add another trigger to pause-client which will activated in case
of OOM as throttling mechanism ([see here](https://github.com/redis/redis/issues/10907)).
In this case we want only to pause client, and eviction actions.
Freeze time during execution of scripts and all other commands.
This means that a key is either expired or not, and doesn't change
state during a script execution. resolves#10182
This PR try to add a new `commandTimeSnapshot` function.
The function logic is extracted from `keyIsExpired`, but the related
calls to `fixed_time_expire` and `mstime()` are removed, see below.
In commands, we will avoid calling `mstime()` multiple times
and just use the one that sampled in call. The background is,
e.g. using `PEXPIRE 1` with valgrind sometimes result in the key
being deleted rather than expired. The reason is that both `PEXPIRE`
command and `checkAlreadyExpired` call `mstime()` separately.
There are other more important changes in this PR:
1. Eliminate `fixed_time_expire`, it is no longer needed.
When we want to sample time we should always use a time snapshot.
We will use `in_nested_call` instead to update the cached time in `call`.
2. Move the call for `updateCachedTime` from `serverCron` to `afterSleep`.
Now `commandTimeSnapshot` will always return the sample time, the
`lookupKeyReadWithFlags` call in `getNodeByQuery` will get a outdated
cached time (because `processCommand` is out of the `call` context).
We put the call to `updateCachedTime` in `aftersleep`.
3. Cache the time each time the module lock Redis.
Call `updateCachedTime` in `moduleGILAfterLock`, affecting `RM_ThreadSafeContextLock`
and `RM_ThreadSafeContextTryLock`
Currently the commandTimeSnapshot change affects the following TTL commands:
- SET EX / SET PX
- EXPIRE / PEXPIRE
- SETEX / PSETEX
- GETEX EX / GETEX PX
- TTL / PTTL
- EXPIRETIME / PEXPIRETIME
- RESTORE key TTL
And other commands just use the cached mstime (including TIME).
This is considered to be a breaking change since it can break a script
that uses a loop to wait for a key to expire.
Fix replication inconsistency on modules that uses key space notifications.
### The Problem
In general, key space notifications are invoked after the command logic was
executed (this is not always the case, we will discuss later about specific
command that do not follow this rules). For example, the `set x 1` will trigger
a `set` notification that will be invoked after the `set` logic was performed, so
if the notification logic will try to fetch `x`, it will see the new data that was written.
Consider the scenario on which the notification logic performs some write
commands. for example, the notification logic increase some counter,
`incr x{counter}`, indicating how many times `x` was changed.
The logical order by which the logic was executed is has follow:
```
set x 1
incr x{counter}
```
The issue is that the `set x 1` command is added to the replication buffer
at the end of the command invocation (specifically after the key space
notification logic was invoked and performed the `incr` command).
The replication/aof sees the commands in the wrong order:
```
incr x{counter}
set x 1
```
In this specific example the order is less important.
But if, for example, the notification would have deleted `x` then we would
end up with primary-replica inconsistency.
### The Solution
Put the command that cause the notification in its rightful place. In the
above example, the `set x 1` command logic was executed before the
notification logic, so it should be added to the replication buffer before
the commands that is invoked by the notification logic. To achieve this,
without a major code refactoring, we save a placeholder in the replication
buffer, when finishing invoking the command logic we check if the command
need to be replicated, and if it does, we use the placeholder to add it to the
replication buffer instead of appending it to the end.
To be efficient and not allocating memory on each command to save the
placeholder, the replication buffer array was modified to reuse memory
(instead of allocating it each time we want to replicate commands).
Also, to avoid saving a placeholder when not needed, we do it only for
WRITE or MAY_REPLICATE commands.
#### Additional Fixes
* Expire and Eviction notifications:
* Expire/Eviction logical order was to first perform the Expire/Eviction
and then the notification logic. The replication buffer got this in the
other way around (first notification effect and then the `del` command).
The PR fixes this issue.
* The notification effect and the `del` command was not wrap with
`multi-exec` (if needed). The PR also fix this issue.
* SPOP command:
* On spop, the `spop` notification was fired before the command logic
was executed. The change in this PR would have cause the replication
order to be change (first `spop` command and then notification `logic`)
although the logical order is first the notification logic and then the
`spop` logic. The right fix would have been to move the notification to
be fired after the command was executed (like all the other commands),
but this can be considered a breaking change. To overcome this, the PR
keeps the current behavior and changes the `spop` code to keep the right
logical order when pushing commands to the replication buffer. Another PR
will follow to fix the SPOP properly and match it to the other command (we
split it to 2 separate PR's so it will be easy to cherry-pick this PR to 7.0 if
we chose to).
#### Unhanded Known Limitations
* key miss event:
* On key miss event, if a module performed some write command on the
event (using `RM_Call`), the `dirty` counter would increase and the read
command that cause the key miss event would be replicated to the replication
and aof. This problem can also happened on a write command that open
some keys but eventually decides not to perform any action. We decided
not to handle this problem on this PR because the solution is complex
and will cause additional risks in case we will want to cherry-pick this PR.
We should decide if we want to handle it in future PR's. For now, modules
writers is advice not to perform any write commands on key miss event.
#### Testing
* We already have tests to cover cases where a notification is invoking write
commands that are also added to the replication buffer, the tests was modified
to verify that the replica gets the command in the correct logical order.
* Test was added to verify that `spop` behavior was kept unchanged.
* Test was added to verify key miss event behave as expected.
* Test was added to verify the changes do not break lazy expiration.
#### Additional Changes
* `propagateNow` function can accept a special dbid, -1, indicating not
to replicate `select`. We use this to replicate `multi/exec` on `propagatePendingCommands`
function. The side effect of this change is that now the `select` command
will appear inside the `multi/exec` block on the replication stream (instead of
outside of the `multi/exec` block). Tests was modified to match this new behavior.
Summary of changes:
1. Rename `redisCommand->name` to `redisCommand->declared_name`, it is a
const char * for native commands and SDS for module commands.
2. Store the [sub]command fullname in `redisCommand->fullname` (sds).
3. List subcommands in `ACL CAT`
4. List subcommands in `COMMAND LIST`
5. `moduleUnregisterCommands` now will also free the module subcommands.
6. RM_GetCurrentCommandName returns full command name
Other changes:
1. Add `addReplyErrorArity` and `addReplyErrorExpireTime`
2. Remove `getFullCommandName` function that now is useless.
3. Some cleanups about `fullname` since now it is SDS.
4. Delete `populateSingleCommand` function from server.h that is useless.
5. Added tests to cover this change.
6. Add some module unload tests and fix the leaks
7. Make error messages uniform, make sure they always contain the full command
name and that it's quoted.
7. Fixes some typos
see the history in #9504, fixes#10124
Co-authored-by: Oran Agra <oran@redislabs.com>
Co-authored-by: guybe7 <guy.benoish@redislabs.com>
The mess:
Some parts use alsoPropagate for late propagation, others using an immediate one (propagate()),
causing edge cases, ugly/hacky code, and the tendency for bugs
The basic idea is that all commands are propagated via alsoPropagate (i.e. added to a list) and the
top-most call() is responsible for going over that list and actually propagating them (and wrapping
them in MULTI/EXEC if there's more than one command). This is done in the new function,
propagatePendingCommands.
Callers to propagatePendingCommands:
1. top-most call() (we want all nested call()s to add to the also_propagate array and just the top-most
one to propagate them) - via `afterCommand`
2. handleClientsBlockedOnKeys: it is out of call() context and it may propagate stuff - via `afterCommand`.
3. handleClientsBlockedOnKeys edge case: if the looked-up key is already expired, we will propagate the
expire but will not unblock any client so `afterCommand` isn't called. in that case, we have to propagate
the deletion explicitly.
4. cron stuff: active-expire and eviction may also propagate stuff
5. modules: the module API allows to propagate stuff from just about anywhere (timers, keyspace notifications,
threads). I could have tried to catch all the out-of-call-context places but it seemed easier to handle it in one
place: when we free the context. in the spirit of what was done in call(), only the top-most freeing of a module
context may cause propagation.
6. modules: when using a thread-safe ctx it's not clear when/if the ctx will be freed. we do know that the module
must lock the GIL before calling RM_Replicate/RM_Call so we propagate the pending commands when
releasing the GIL.
A "known limitation", which were actually a bug, was fixed because of this commit (see propagate.tcl):
When using a mix of RM_Call with `!` and RM_Replicate, the command would propagate out-of-order:
first all the commands from RM_Call, and then the ones from RM_Replicate
Another thing worth mentioning is that if, in the past, a client would issue a MULTI/EXEC with just one
write command the server would blindly propagate the MULTI/EXEC too, even though it's redundant.
not anymore.
This commit renames propagate() to propagateNow() in order to cause conflicts in pending PRs.
propagatePendingCommands is the only caller of propagateNow, which is now a static, internal helper function.
Optimizations:
1. alsoPropagate will not add stuff to also_propagate if there's no AOF and replicas
2. alsoPropagate reallocs also_propagagte exponentially, to save calls to memmove
Bugfixes:
1. CONFIG SET can create evictions, sending notifications which can cause to dirty++ with modules.
we need to prevent it from propagating to AOF/replicas
2. We need to set current_client in RM_Call. buggy scenario:
- CONFIG SET maxmemory, eviction notifications, module hook calls RM_Call
- assertion in lookupKey crashes, because current_client has CONFIG SET, which isn't CMD_WRITE
3. minor: in eviction, call propagateDeletion after notification, like active-expire and all commands
(we always send a notification before propagating the command)
- Added sanitizer support. `address`, `undefined` and `thread` sanitizers are available.
- To build Redis with desired sanitizer : `make SANITIZER=undefined`
- There were some sanitizer findings, cleaned up codebase
- Added tests with address and undefined behavior sanitizers to daily CI.
- Added tests with address sanitizer to the per-PR CI (smoke out mem leaks sooner).
Basically, there are three types of issues :
**1- Unaligned load/store** : Most probably, this issue may cause a crash on a platform that
does not support unaligned access. Redis does unaligned access only on supported platforms.
**2- Signed integer overflow.** Although, signed overflow issue can be problematic time to time
and change how compiler generates code, current findings mostly about signed shift or simple
addition overflow. For most platforms Redis can be compiled for, this wouldn't cause any issue
as far as I can tell (checked generated code on godbolt.org).
**3 -Minor leak** (redis-cli), **use-after-free**(just before calling exit());
UB means nothing guaranteed and risky to reason about program behavior but I don't think any
of the fixes here worth backporting. As sanitizers are now part of the CI, preventing new issues
will be the real benefit.
Reduce dict struct memory overhead
on 64bit dict size goes down from jemalloc's 96 byte bin to its 56 byte bin.
summary of changes:
- Remove `privdata` from callbacks and dict creation. (this affects many files, see "Interface change" below).
- Meld `dictht` struct into the `dict` struct to eliminate struct padding. (this affects just dict.c and defrag.c)
- Eliminate the `sizemask` field, can be calculated from size when needed.
- Convert the `size` field into `size_exp` (exponent), utilizes one byte instead of 8.
Interface change: pass dict pointer to dict type call back functions.
This is instead of passing the removed privdata field. In the future if
we'd like to have private data in the callbacks we can extract it from
the dict type. We can extend dictType to include a custom dict struct
allocator and use it to allocate more data at the end of the dict
struct. This data can then be used to store private data later acccessed
by the callbacks.
Add NX, XX, GT, and LT flags to EXPIRE, PEXPIRE, EXPIREAT, PEXAPIREAT.
- NX - only modify the TTL if no TTL is currently set
- XX - only modify the TTL if there is a TTL currently set
- GT - only increase the TTL (considering non-volatile keys as infinite expire time)
- LT - only decrease the TTL (considering non-volatile keys as infinite expire time)
return value of the command is 0 when the operation was skipped due to one of these flags.
Signed-off-by: Ning Sun <sunng@protonmail.com>
The root cause is that one test (`5 keys in, 5 keys out`) is leaking a volatile key
that can expire while another later test(`All TTL in commands are propagated
as absolute timestamp in replication stream`) is running.
Such leaked expiration injects an unexpected `DEL` command into the
replication command during the later test, causing it to fail.
The fixes are two fold:
1. Plug the leak in the first test.
2. Add FLUSHALL to the later test, to avoid future interference from other tests.
Till now, on replica full-sync we used to transfer absolute time for TTL,
however when a command arrived (EXPIRE or EXPIREAT),
we used to propagate it as is to replicas (possibly with relative time),
but always translate it to EXPIREAT (absolute time) to AOF.
This commit changes that and will always use absolute time for propagation.
see discussion in #8433
Furthermore, we Introduce new commands: `EXPIRETIME/PEXPIRETIME`
that allow extracting the absolute TTL time from a key.
Respond with error if expire time overflows from positive to negative of vice versa.
* `SETEX`, `SET EX`, `GETEX` etc would have already error on negative value,
but now they would also error on overflows (i.e. when the input was positive but
after the manipulation it becomes negative, which would have passed before)
* `EXPIRE` and `EXPIREAT` was ok taking negative values (would implicitly delete
the key), we keep that, but we do error if the user provided a value that changes
sign when manipulated (except the case of changing sign when `basetime` is added)
Signed-off-by: Gnanesh <gnaneshkunal@outlook.com>
Co-authored-by: Oran Agra <oran@redislabs.com>
As we know, redis may reject user's requests or evict some keys if
used memory is over maxmemory. Dictionaries expanding may make
things worse, some big dictionaries, such as main db and expires dict,
may eat huge memory at once for allocating a new big hash table and be
far more than maxmemory after expanding.
There are related issues: #4213#4583
More details, when expand dict in redis, we will allocate a new big
ht[1] that generally is double of ht[0], The size of ht[1] will be
very big if ht[0] already is big. For db dict, if we have more than
64 million keys, we need to cost 1GB for ht[1] when dict expands.
If the sum of used memory and new hash table of dict needed exceeds
maxmemory, we shouldn't allow the dict to expand. Because, if we
enable keys eviction, we still couldn't add much more keys after
eviction and rehashing, what's worse, redis will keep less keys when
redis only remains a little memory for storing new hash table instead
of users' data. Moreover users can't write data in redis if disable
keys eviction.
What this commit changed ?
Add a new member function expandAllowed for dict type, it provide a way
for caller to allow expand or not. We expose two parameters for this
function: more memory needed for expanding and dict current load factor,
users can implement a function to make a decision by them.
For main db dict and expires dict type, these dictionaries may be very
big and cost huge memory for expanding, so we implement a judgement
function: we can stop dict to expand provisionally if used memory will
be over maxmemory after dict expands, but to guarantee the performance
of redis, we still allow dict to expand if dict load factor exceeds the
safe load factor.
Add test cases to verify we don't allow main db to expand when left
memory is not enough, so that avoid keys eviction.
Other changes:
For new hash table size when expand. Before this commit, the size is
that double used of dict and later _dictNextPower. Actually we aim to
control a dict load factor between 0.5 and 1.0. Now we replace *2 with
+1, since the first check is that used >= size, the outcome of before
will usually be the same as _dictNextPower(used+1). The only case where
it'll differ is when dict_can_resize is false during fork, so that later
the _dictNextPower(used*2) will cause the dict to jump to *4 (i.e.
_dictNextPower(1025*2) will return 4096).
Fix rehash test cases due to changing algorithm of new hash table size
when expand.
This wrong behavior was backed by a test, and also documentation, and dates back to 2010.
But it makes no sense to anyone involved so it was decided to change that.
Note that 20eeddf (invalidate watch on expire on access) was released in 6.0 RC2
and 2d1968f released in in 6.0.0 GA (invalidate watch when key is evicted).
both of which do similar changes.
Similarly to EXPIREAT with TTL in the past, which implicitly deletes the
key and return success, RESTORE should not store key that are already
expired into the db.
When used together with REPLACE it should emit a DEL to keyspace
notification and replication stream.
Likely fix#6723.
This is what happens AFAIK: we enter the main loop where we expire stuff
until a given percentage of keys is still found to be logically expired.
There are however other potential exit conditions.
However the "sampled" variable is not always incremented inside the
loop, because we may found no valid slot as we scan the hash table, but
just NULLs ad dict entries. So when the do/while loop condition is
triggered at the end, we do (expired*100/sampled), dividing by zero if
we sampled 0 keys.
Otherwise what happens is that the tracking table will never get garbage
collected if there are no longer clients with tracking enabled.
Now the invalidation function immediately checks if there is any table
allocated, otherwise it returns ASAP, so the overhead when the feature
is not used should be near zero.
Pieter Cailliau