Now both master and replicas keep track of the last replication offset
that contains meaningful data (ignoring the tailing pings), and both
trim that tail from the replication backlog, and the offset with which
they try to use for psync.
the implication is that if someone missed some pings, or even have
excessive pings that the promoted replica has, it'll still be able to
psync (avoid full sync).
the downside (which was already committed) is that replicas running old
code may fail to psync, since the promoted replica trims pings form it's
backlog.
This commit adds a test that reproduces several cases of promotions and
demotions with stale and non-stale pings
Background:
The mearningful offset on the master was added recently to solve a problem were
the master is left all alone, injecting PINGs into it's backlog when no one is
listening and then gets demoted and tries to replicate from a replica that didn't
have any of the PINGs (or at least not the last ones).
however, consider this case:
master A has two replicas (B and C) replicating directly from it.
there's no traffic at all, and also no network issues, just many pings in the
tail of the backlog. now B gets promoted, A becomes a replica of B, and C
remains a replica of A. when A gets demoted, it trims the pings from its
backlog, and successfully replicate from B. however, C is still aware of
these PINGs, when it'll disconnect and re-connect to A, it'll ask for something
that's not in the backlog anymore (since A trimmed the tail of it's backlog),
and be forced to do a full sync (something it didn't have to do before the
meaningful offset fix).
Besides that, the psync2 test was always failing randomly here and there, it
turns out the reason were PINGs. Investigating it shows the following scenario:
cycle 1: redis #1 is master, and all the rest are direct replicas of #1
cycle 2: redis #2 is promoted to master, #1 is a replica of #2 and #3 is replica of #1
now we see that when #1 is demoted it prints:
17339:S 21 Apr 2020 11:16:38.523 * Using the meaningful offset 3929963 instead of 3929977 to exclude the final PINGs (14 bytes difference)
17339:S 21 Apr 2020 11:16:39.391 * Trying a partial resynchronization (request e2b3f8817735fdfe5fa4626766daa938b61419e5:3929964).
17339:S 21 Apr 2020 11:16:39.392 * Successful partial resynchronization with master.
and when #3 connects to the demoted #2, #2 says:
17339:S 21 Apr 2020 11:16:40.084 * Partial resynchronization not accepted: Requested offset for secondary ID was 3929978, but I can reply up to 3929964
so the issue here is that the meaningful offset feature saved the day for the
demoted master (since it needs to sync from a replica that didn't get the last
ping), but it didn't help one of the other replicas which did get the last ping.
STRALGO should be a container for mostly read-only string
algorithms in Redis. The algorithms should have two main
characteristics:
1. They should be non trivial to compute, and often not part of
programming language standard libraries.
2. They should be fast enough that it is a good idea to have optimized C
implementations.
Next thing I would love to see? A small strings compression algorithm.
Reloading of the RDB generated by
DEBUG POPULATE 5000000
SAVE
is now 25% faster.
This commit also prepares the ability to have more flexibility when
loading stuff from the RDB, since we no longer use dbAdd() but can
control exactly how things are added in the database.
Related to #5145.
Design note: clients may change type when they turn into replicas or are
moved into the Pub/Sub category and so forth. Moreover the recomputation
of the bytes used is problematic for obvious reasons: it changes
continuously, so as a conservative way to avoid accumulating errors,
each client remembers the contribution it gave to the sum, and removes
it when it is freed or before updating it with the new memory usage.
Makse sure call() doesn't wrap replicated commands with
a redundant MULTI/EXEC
Other, unrelated changes:
1. Formatting compiler warning in INFO CLIENTS
2. Use CLIENT_ID_AOF instead of UINT64_MAX
37a10cef introduced automatic wrapping of MULTI/EXEC for the
alsoPropagate API. However this collides with the built-in mechanism
already present in module.c. To avoid complex changes near Redis 6 GA
this commit introduces the ability to exclude call() MUTLI/EXEC wrapping
for also propagate in order to continue to use the old code paths in
module.c.
Now that this mechanism is the sole one used for blocked clients
timeouts, it is more wise to cleanup the table when the client unblocks
for any reason. We use a flag: CLIENT_IN_TO_TABLE, in order to avoid a
radix tree lookup when the client was already removed from the table
because we processed it by scanning the radix tree.
A very commonly signaled operational problem with Redis master-replicas
sets is that, once the master becomes unavailable for some reason,
especially because of network problems, many times it wont be able to
perform a partial resynchronization with the new master, once it rejoins
the partition, for the following reason:
1. The master becomes isolated, however it keeps sending PINGs to the
replicas. Such PINGs will never be received since the link connection is
actually already severed.
2. On the other side, one of the replicas will turn into the new master,
setting its secondary replication ID offset to the one of the last
command received from the old master: this offset will not include the
PINGs sent by the master once the link was already disconnected.
3. When the master rejoins the partion and is turned into a replica, its
offset will be too advanced because of the PINGs, so a PSYNC will fail,
and a full synchronization will be required.
Related to issue #7002 and other discussion we had in the past around
this problem.
Before this commit, when upgrading a replica, expired keys will not
be loaded, thus causing replica having less keys in db. To this point,
master and replica's keys is logically consistent. However, before
the keys in master and replica are physically consistent, that is,
they have the same dbsize, if master got a problem and the replica
got promoted and becomes new master of that partition, and master
updates a key which does not exist on master, but physically exists
on the old master(new replica), the old master would refuse to update
the key, thus causing master and replica data inconsistent.
How could this happen?
That's all because of the wrong judgement of roles while starting up
the server. We can not use server.masterhost to judge if the server
is master or replica, since it fails in cluster mode.
When we start the server, we load rdb and do want to load expired keys,
and do not want to have the ability to active expire keys, if it is
a replica.
1. Call emptyDb even in case of diskless-load: We want modules
to get the same FLUSHDB event as disk-based replication.
2. Do not fire any module events when flushing the backups array.
3. Delete redundant call to signalFlushedDb (Called from emptyDb).
Because "keymiss" is "special" compared to the rest of
the notifications (Trying not to break existing apps
using the 'A' format for notifications)
Also updated redis.conf and module.c docs
Checking OOM by `getMaxMemoryState` inside script might get different result
with `freeMemoryIfNeededAndSafe` at script start, because lua stack and
arguments also consume memory.
This leads to memory `borderline` when memory grows near server.maxmemory:
- `freeMemoryIfNeededAndSafe` at script start detects no OOM, no memory freed
- `getMaxMemoryState` inside script detects OOM, script aborted
We solve this 'borderline' issue by saving OOM state at script start to get
stable lua OOM state.
related to issue #6565 and #5250.
