Turns out the RDB checksum in Redis 6.0 on bigendian is broken.
It always returned 0, so the RDB files are generated as if checksum is
disabled, and will be loaded ok on littleendian, and on bigendian.
But it'll not be able to load RDB files generated on littleendian or older versions.
Similarly DUMP and RESTORE will work on the same version (0==0),
but will be unable to exchange dump payloads with littleendian or old versions.
When a Lua script returns a map to redis (a feature which was added in
redis 6 together with RESP3), it would have returned the value first and
the key second.
If the client was using RESP2, it was getting them out of order, and if
the client was in RESP3, it was getting a map of value => key.
This was happening regardless of the Lua script using redis.setresp(3)
or not.
This also affects a case where the script was returning a map which it got
from from redis by doing something like: redis.setresp(3); return redis.call()
This fix is a breaking change for redis 6.0 users who happened to rely
on the wrong order (either ones that used redis.setresp(3), or ones that
returned a map explicitly).
This commit also includes other two changes in the tests:
1. The test suite now handles RESP3 maps as dicts rather than nested
lists
2. Remove some redundant (duplicate) tests from tracking.tcl
* man-like consistent long formatting
* Uppercases commands, subcommands and options
* Adds 'HELP' to HELP for all
* Lexicographical order
* Uses value notation and other .md likeness
* Moves const char *help to top
* Keeps it under 80 chars
* Misc help typos, consistent conjuctioning (i.e return and not returns)
* Uses addReplySubcommandSyntaxError(c) all over
Signed-off-by: Itamar Haber <itamar@redislabs.com>
This PR not only fixes the problem that swapdb does not make the
transaction fail, but also optimizes the FLUSHALL and FLUSHDB command to
set the CLIENT_DIRTY_CAS flag to avoid unnecessary traversal of clients.
FLUSHDB was changed to first iterate on all watched keys, and then on the
clients watching each key.
Instead of iterating though all clients, and for each iterate on watched keys.
Co-authored-by: Oran Agra <oran@redislabs.com>
If AOF file contains a long Lua script that timed out, then the `evalCommand` calls
`blockingOperationEnds` which sets `server.blocked_last_cron` to 0. later on,
the AOF `whileBlockedCron` function asserts that this value is not 0.
The fix allows nesting call to `blockingOperationStarts` and `blockingOperationEnds`.
The issue was first introduce in this commit: 9ef8d2f67 (Redis 6.2 RC1)
This is a recent problem, introduced by 7471743 (redis 6.0)
The implications are:
The sole difference between LookupKeyRead and LookupKeyWrite is for command
executed on a replica, which are not received from its master client. (for the master,
and for the master client on the replica, these two functions behave the same)!
Since SORT is a write command, this bug only implicates a writable-replica.
And these are its implications:
- SORT STORE will behave as it did before the above mentioned commit (like before
redis 6.0). on a writable-replica an already logically expired the key would have
appeared missing. (store dest key would be deleted, instead of being populated
with the data from the already logically expired key)
- SORT (the non store variant, which in theory could have been executed on
read-only-replica if it weren't for the write flag), will (in redis 6.0) have a new bug
and return the data from the already logically expired key instead of empty response.
New command flags similar to what SADD already has.
Co-authored-by: huangwei03 <huangwei03@kuaishou.com>
Co-authored-by: Itamar Haber <itamar@redislabs.com>
Co-authored-by: Oran Agra <oran@redislabs.com>
In #7726 (part of 6.2), we added a mechanism for whileBlockedCron, this
mechanism has an assertion to make sure the timestamp in
whileBlockedCron was always set correctly before the blocking operation
starts.
I now found (thanks to our CI) two bugs in that area:
1) CONFIG RESETSTAT (if it was allowed during loading) would have
cleared this var
2) the call stopLoading (which calls whileBlockedCron) was made too
early, while the rio is still in use, in which case the update_cksum
(rdbLoadProgressCallback) may still be called and whileBlockedCron
can assert.
getRDB is "designed" to work in two modes: one for redis-cli --rdb and
one for redis-cli --cluster backup.
in the later case it uses the hiredis connection from the cluster nodes
and it used to free it without nullifying the context, so a later
attempt to free the context would crash.
I suppose the reason it seems to want to free the hiredis context ASAP
is that it wants to disconnect the replica link, so that replication
buffers will not be accumulated.
The crash log attempts to print the current client info, and when it
does that it attempts to check if the first argument happens to be a key
but it did so for commands with no arguments too, which caused the crash
log to crash half way and not reach its end.
This Commit pushes forward the observability on overall error statistics and command statistics within redis-server:
It extends INFO COMMANDSTATS to have
- failed_calls in - so we can keep track of errors that happen from the command itself, broken by command.
- rejected_calls - so we can keep track of errors that were triggered outside the commmand processing per se
Adds a new section to INFO, named ERRORSTATS that enables keeping track of the different errors that
occur within redis ( within processCommand and call ) based on the reply Error Prefix ( The first word
after the "-", up to the first space ).
This commit also fixes RM_ReplyWithError so that it can be correctly identified as an error reply.
Recently efaf09ee4 started using addReplyErrorSds in place of
addReplySds the later takes ownership of the string but the former did
not.
This introduced memory leaks when a script returns an error to redis,
and also in clusterRedirectClient (two new usages of
addReplyErrorSds which was mostly unused till now.
This commit chagnes two thanks.
1. change addReplyErrorSds to take ownership of the error string.
2. scripting.c doesn't actually need to use addReplyErrorSds, it's a
perfect match for addReplyErrorFormat (replaces newlines with spaces)
Adds: `L/RPOP <key> [count]`
Implements no. 2 of the following strategies:
1. Loop on listTypePop - this would result in multiple calls for memory freeing and allocating (see 769167a079)
2. Iterate the range to build the reply, then call quickListDelRange - this requires two iterations and **is the current choice**
3. Refactor quicklist to have a pop variant of quickListDelRange - probably optimal but more complex
Also:
* There's a historical check for NULL after calling listTypePop that was converted to an assert.
* This refactors common logic shared between LRANGE and the new form of LPOP/RPOP into addListRangeReply (adds test for b/w compat)
* Consequently, it may have made sense to have `LRANGE l -1 -2` and `LRANGE l 9 0` be legit and return a reverse reply. Due to historical reasons that would be, however, a breaking change.
* Added minimal comments to existing commands to adhere to the style, make core dev life easier and get commit karma, naturally.
The commit deals with the syncWithMaster and the ugly state machine in it.
It attempts to clean it a bit, but more importantly it uses pipeline for
part of the work (rather than 7 round trips, we now have 4).
i.e. the connect and PING are separate, then AUTH + 3 REPLCONF in one pipeline,
and finally the PSYNC (must be separate since the master has to have an empty
output buffer).
This is just a refactoring commit.
This function was never actually used as a synchronous (do both send or
receive), it was always used only ine one of the two modes, which meant it
has to take extra arguments that are not relevant for the other.
Besides that, a tool that sends a synchronous command, it not something
we want in our toolbox (synchronous IO in single threaded app is evil).
sendSynchronousCommand was now refactored into separate sending and
receiving APIs, and the sending part has two variants, one taking vaargs,
and the other taking argc+argv (and an optional length array which means
you can use binary sds strings).
As discussed in https://github.com/antirez/redis/issues/7364, it is good
to have a HELLO command variant, which does not switch the current proto
version of a redis server.
While `HELLO` will work, it introduced a certain difficulty on parsing
options of the command. We will need to offset the index of authentication
and setname option by -1.
So 0 is marked a special version meaning non-switching. And we do not
need to change the code much.
When a database on a 64 bit build grows past 2^31 keys, the underlying hash table expands to 2^32 buckets. After this point, the algorithms for selecting random elements only return elements from half of the available buckets because they use random() which has a range of 0 to 2^31 - 1. This causes problems for eviction policies which use dictGetSomeKeys or dictGetRandomKey. Over time they cause the hash table to become unbalanced because, while new keys are spread out evenly across all buckets, evictions come from only half of the available buckets. Eventually this half of the table starts to run out of keys and it takes longer and longer to find candidates for eviction. This continues until no more evictions can happen.
This solution addresses this by using a 64 bit PRNG instead of libc random().
Co-authored-by: Greg Femec <gfemec@google.com>
Turns out that when the fork child crashes, the crash log was deleting
the pidfile from the disk (although the parent is still running.
Now we set the pidfile of the fork process to NULL so the fork process
will never deletes it.
Normally IO threads should simply read data from the socket into the
buffer and attempt to parse it.
If a protocol error is detected, a reply is generated which may result
with installing a write handler which is not thread safe. This fix
delays that until the client is processed back in the main thread.
Fixes#8220
In the distant history there was only the read flag for commands, and whatever
command that didn't have the read flag was a write one.
Then we added the write flag, but some portions of the code still used !read
Also some commands that don't work on the keyspace at all, still have the read
flag.
Changes in this commit:
1. remove the read-only flag from TIME, ECHO, ROLE and LASTSAVE
2. EXEC command used to decides if it should propagate a MULTI by looking at
the command flags (!read & !admin).
When i was about to change it to look at the write flag instead, i realized
that this would cause it not to propagate a MULTI for PUBLISH, EVAL, and
SCRIPT, all 3 are not marked as either a read command or a write one (as
they should), but all 3 are calling forceCommandPropagation.
So instead of introducing a new flag to denote a command that "writes" but
not into the keyspace, and still needs propagation, i decided to rely on
the forceCommandPropagation, and just fix the code to propagate MULTI when
needed rather than depending on the command flags at all.
The implication of my change then is that now it won't decide to propagate
MULTI when it sees one of these: SELECT, PING, INFO, COMMAND, TIME and
other commands which are neither read nor write.
3. Changing getNodeByQuery and clusterRedirectBlockedClientIfNeeded in
cluster.c to look at !write rather than read flag.
This should have no implications, since these code paths are only reachable
for commands which access keys, and these are always marked as either read
or write.
This commit improve MULTI propagation tests, for modules and a bunch of
other special cases, all of which used to pass already before that commit.
the only one that test change that uncovered a change of behavior is the
one that DELs a non-existing key, it used to propagate an empty
multi-exec block, and no longer does.
In response to large client query buffer optimization introduced in 1898e6c. The calculation of the amount of
remaining bytes we need to write to the query buffer was calculated wrong, as a result we are unnecessarily
growing the client query buffer by sdslen(c->querybuf) always. This fix corrects that behavior.
Please note the previous behavior prior to the before-mentioned change was correctly calculating the remaining
additional bytes, and this change makes that calculate to be consistent.
Useful context, the argument of size `ll` starts at qb_pos (which is now the beginning of the sds), but much of it
may have already been read from the socket, so we only need to grow the sds for the remainder of it.
If we only has one node in cluster or before 8fdc857, we don't know myself ip, so we should use config.hostip for myself.
However, we should use the IP from the command response to update node->ip if it exists and is different from config.hostip
otherwise, when there's more than one node in cluster, if we use -h with virtual IP or DNS, benchmark doesn't show node real ip and port of myself even though it could get right IP and port by CLUSTER NODES command.
Fix wrong server dirty increment in
* spopWithCountCommand
* hsetCommand
* ltrimCommand
* pfaddCommand
Some didn't increment the amount of fields (just one per command).
Others had excessive increments.
Exposes the main thread CPU info via info modules ( linux specific only )
(used_cpu_sys_main_thread and used_cpu_user_main_thread). This is important for:
- distinguish between main thread and io-threads cpu time total cpu time consumed ( check
what is the first bottleneck on the used config )
- distinguish between main thread and modules threads total cpu time consumed
Apart from it, this commit also exposes the server_time_usec within the Server section so that we can
properly differentiate consecutive collection and calculate for example the CPU% and or / cpu time vs
wall time, etc...
* Allow runtest-moduleapi use a different 'make', for systems where GNU Make is 'gmake'.
* Fix issue with builds on Solaris re-building everything from scratch due to CFLAGS/LDFLAGS not stored.
* Fix compile failure on Solaris due to atomicvar and a bunch of warnings.
* Fix garbled log timestamps on Solaris.
When a replica uses the diskless-load swapdb approach, it backs up the old database,
then attempts to load a new one, and in case of failure, it restores the backup.
this means that modules with global out of keyspace data, must have an option to
subscribe to events and backup/restore/discard their global data too.
Add a new set of defrag functions that take a defrag context and allow
defragmenting memory blocks and RedisModuleStrings.
Modules can register a defrag callback which will be invoked when the
defrag process handles globals.
Modules with custom data types can also register a datatype-specific
defrag callback which is invoked for keys that require defragmentation.
The callback and associated functions support both one-step and
multi-step options, depending on the complexity of the key as exposed by
the free_effort callback.
The pid of the benchmark process is used to randomize the random number generator's
seed. This ensures that when multiple benchmark processes are started at the same time
to generate load on a server, they use different seeds. This will ensure randomness in
the keys generated by different benchmark processes.
Add commands to query geospatial data with bounding box.
Two new commands that replace the existing 4 GEORADIUS* commands.
GEOSEARCH key [FROMMEMBER member] [FROMLOC long lat] [BYRADIUS radius
unit] [BYBOX width height unit] [WITHCORD] [WITHDIST] [WITHASH] [COUNT
count] [ASC|DESC]
GEOSEARCHSTORE dest_key src_key [FROMMEMBER member] [FROMLOC long lat]
[BYRADIUS radius unit] [BYBOX width height unit] [WITHCORD] [WITHDIST]
[WITHASH] [COUNT count] [ASC|DESC] [STOREDIST]
- Add two types of CIRCULAR_TYPE and RECTANGLE_TYPE to achieve different searches
- Judge whether the point is within the rectangle, refer to:
geohashGetDistanceIfInRectangle
This adds a new `tls-client-cert-file` and `tls-client-key-file`
configuration directives which make it possible to use different
certificates for the TLS-server and TLS-client functions of Redis.
This is an optional directive. If it is not specified the `tls-cert-file`
and `tls-key-file` directives are used for TLS client functions as well.
Also, `utils/gen-test-certs.sh` now creates additional server-only and client-only certs and will skip intensive operations if target files already exist.
This adds a copy callback for module data types, in order to make
modules compatible with the new COPY command.
The callback is optional and COPY will fail for keys with data types
that do not implement it.
Module blocked clients cache the response in a temporary client,
the reply list in this client would be affected by the recent fix
in #7202, but when the reply is later copied into the real client,
it would have bypassed all the checks for output buffer limit, which
would have resulted in both: responding with a partial response to
the client, and also not disconnecting it at all.
Makes spt_init more careful with assumptions about what memory regions
may be overwritten. It will now only consider a contiguous block of argv
and envp elements and mind any gaps.
when the same consumer re-claim an entry that it already has, there's
no need to remove-and-insert if it's the same rax.
we do need to update the idle time though.
this commit only improves efficiency (doesn't change behavior).
* Add CLIENT INFO subcommand.
The output is identical to CLIENT LIST but provides a single line for
the current client only.
* Add CLIENT LIST ID [id...].
Co-authored-by: Itamar Haber <itamar@redislabs.com>
First, if the ziplist header is surely inside the ziplist, do fast path
decoding rather than the careful one.
In that case, streamline the encoding if-else chain to be executed only
once, and the encoding validity tested at the end.
encourage inlining
likely / unlikely hints for speculative execution
Assertion used _exit(1) to tell the compiler that the code after them is
not reachable and get rid of warnings.
But in some cases assertions are placed inside tight loops, and any
piece of code in them can slow down execution (code cache and other
reasons), instead using either abort() or better yet, unreachable
builtin.
If RESTORE passes successfully with full sanitization, we can't affort
to crash later on assertion due to duplicate records in a hash when
converting it form ziplist to dict.
This means that when doing full sanitization, we must make sure there
are no duplicate records in any of the collections.
The test creates keys with various encodings, DUMP them, corrupt the payload
and RESTORES it.
It utilizes the recently added use-exit-on-panic config to distinguish between
asserts and segfaults.
If the restore succeeds, it runs random commands on the key to attempt to
trigger a crash.
It runs in two modes, one with deep sanitation enabled and one without.
In the first one we don't expect any assertions or segfaults, in the second one
we expect assertions, but no segfaults.
We also check for leaks and invalid reads using valgrind, and if we find them
we print the commands that lead to that issue.
Changes in the code (other than the test):
- Replace a few NPD (null pointer deference) flows and division by zero with an
assertion, so that it doesn't fail the test. (since we set the server to use
`exit` rather than `abort` on assertion).
- Fix quite a lot of flows in rdb.c that could have lead to memory leaks in
RESTORE command (since it now responds with an error rather than panic)
- Add a DEBUG flag for SET-SKIP-CHECKSUM-VALIDATION so that the test don't need
to bother with faking a valid checksum
- Remove a pile of code in serverLogObjectDebugInfo which is actually unsafe to
run in the crash report (see comments in the code)
- fix a missing boundary check in lzf_decompress
test suite infra improvements:
- be able to run valgrind checks before the process terminates
- rotate log files when restarting servers
When loading an encoded payload we will at least do a shallow validation to
check that the size that's encoded in the payload matches the size of the
allocation.
This let's us later use this encoded size to make sure the various offsets
inside encoded payload don't reach outside the allocation, if they do, we'll
assert/panic, but at least we won't segfault or smear memory.
We can also do 'deep' validation which runs on all the records of the encoded
payload and validates that they don't contain invalid offsets. This lets us
detect corruptions early and reject a RESTORE command rather than accepting
it and asserting (crashing) later when accessing that payload via some command.
configuration:
- adding ACL flag skip-sanitize-payload
- adding config sanitize-dump-payload [yes/no/clients]
For now, we don't have a good way to ensure MIGRATE in cluster resharding isn't
being slowed down by these sanitation, so i'm setting the default value to `no`,
but later on it should be set to `clients` by default.
changes:
- changing rdbReportError not to `exit` in RESTORE command
- adding a new stat to be able to later check if cluster MIGRATE isn't being
slowed down by sanitation.
When client tracking is enabled signalModifiedKey can increase memory usage,
this can cause the loop in performEvictions to keep running since it was measuring
the memory usage impact of signalModifiedKey.
The section that measures the memory impact of the eviction should be just on dbDelete,
excluding keyspace notification, client tracking, and propagation to AOF and replicas.
This resolves part of the problem described in #8069
p.s. fix took 1 minute, test took about 3 hours to write.
One way this was happening is when a module issued an RM_Call which would inject MULTI.
If the module command that does that was itself issued by something else that already did
added MULTI (e.g. another module, or a Lua script), it would have caused nested MULTI.
In fact the MULTI state in the client or the MULTI_EMITTED flag in the context isn't
the right indication that we need to propagate MULTI or not, because on a nested calls
(possibly a module action called by a keyspace event of another module action), these
flags aren't retained / reflected.
instead there's now a global propagate_in_transaction flag for that.
in addition to that, we now have a global in_eval and in_exec flags, to serve the flags
of RM_GetContextFlags, since their dependence on the current client is wrong for the same
reasons mentioned above.
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.
In the iterator for these functions, we'll traverse the sorted sets
in a reversed way so that largest elements come first. We prefer
this order because it's optimized for insertion in a skiplist, which
is the destination of the elements being iterated in there functions.
When replica diskless-load type is swapdb in cluster mode, we didn't backup
keys to slots map, so we will lose keys to slots map if fail to sync.
Now we backup keys to slots map at first, and restore it properly when fail.
This commit includes a refactory/cleanup of the backups mechanism (moving it to db.c and re-structuring it a bit).
Co-authored-by: Oran Agra <oran@redislabs.com>
As described in redis-benchamrk help message 'The test names are the same as the ones produced as output.', In redis-benchmark output, we can only see PING_BULK, but the cmd `redis-benchmark -t ping_bulk` is not supported. We have to run it with ping_mbulk which is not user friendly.
SELECT used to read the index into a `long` variable, and then pass it to a function
that takes an `int`, possibly causing an overflow before the range check.
Now all these commands use better and cleaner range check, and that also results in
a slight change of the error response in case of an invalid database index.
SELECT:
in the past it would have returned either `-ERR invalid DB index` (if not a number),
or `-ERR DB index is out of range` (if not between 1..16 or alike).
now it'll return either `-ERR value is out of range` (if not a number), or
`-ERR value is out of range, value must between -2147483648 and 2147483647`
(if not in the range for an int), or `-ERR DB index is out of range`
(if not between 0..16 or alike)
MOVE:
in the past it would only fail with `-ERR index out of range` no matter the reason.
now return the same errors as the new ones for SELECT mentioned above.
(i.e. unlike for SELECT even for a value like 17 we changed the error message)
COPY:
doesn't really matter how it behaved in the past (new command), new behavior is
like the above two.
Fixes#7923.
This PR appropriates the special `&` symbol (because `@` and `*` are taken),
followed by a literal value or pattern for describing the Pub/Sub patterns that
an ACL user can interact with. It is similar to the existing key patterns
mechanism in function (additive) and implementation (copy-pasta). It also adds
the allchannels and resetchannels ACL keywords, naturally.
The default user is given allchannels permissions, whereas new users get
whatever is defined by the acl-pubsub-default configuration directive. For
backward compatibility in 6.2, the default of this directive is allchannels but
this is likely to be changed to resetchannels in the next major version for
stronger default security settings.
Unless allchannels is set for the user, channel access permissions are checked
as follows :
* Calls to both PUBLISH and SUBSCRIBE will fail unless a pattern matching the
argumentative channel name(s) exists for the user.
* Calls to PSUBSCRIBE will fail unless the pattern(s) provided as an argument
literally exist(s) in the user's list.
Such failures are logged to the ACL log.
Runtime changes to channel permissions for a user with existing subscribing
clients cause said clients to disconnect unless the new permissions permit the
connections to continue. Note, however, that PSUBSCRIBErs' patterns are matched
literally, so given the change bar:* -> b*, pattern subscribers to bar:* will be
disconnected.
Notes/questions:
* UNSUBSCRIBE, PUNSUBSCRIBE and PUBSUB remain unprotected due to lack of reasons
for touching them.
