Security caches the result of role lookups and negative lookups are
cached indefinitely. In the case of transient failures this leads to a
bad experience as the roles could truly exist. The CompositeRolesStore
needs to know if a failure occurred in one of the roles stores in order
to make the appropriate decision as it relates to caching. In order to
provide this information to the CompositeRolesStore, the return type of
methods to retrieve roles has changed to a new class,
RoleRetrievalResult. This class provides the ability to pass back an
exception to the roles store. This exception does not mean that a
request should be failed but instead serves as a signal to the roles
store that missing roles should not be cached and neither should the
combined role if there are missing roles.
As part of this, the negative lookup cache was also changed from an
unbounded cache to a cache with a configurable limit.
Relates #33205
PR #34290 made it impossible to use thread-context values to pass
authentication metadata out of a realm. The SAML realm used this
technique to allow the SamlAuthenticateAction to process the parsed
SAML token, and apply them to the access token that was generated.
This new method adds metadata to the AuthenticationResult itself, and
then the authentication service makes this result available on the
thread context.
Closes: #34332
The ingest pipeline that is produced is very simple. It
contains a grok processor if the format is semi-structured
text, a date processor if the format contains a timestamp,
and a remove processor if required to remove the interim
timestamp field parsed out of semi-structured text.
Eventually the UI should offer the option to customize the
pipeline with additional processors to perform other data
preparation steps before ingesting data to an index.
In ccb9ab5717 we changed how we deal with time
fields to support the `DateTime`-format fields added in 6.0, but dropped
support for pre-6.x `Long`-format fields. This change reinstates this support
for cases where pre-6.x data is made available to ML (e.g. in a mixed-version
CCS setup or after an upgrade).
ListenableFuture may run a listener on the same thread that called the
addListener method or it may execute on another thread after the future
has completed. Whenever the ListenableFuture stores the listener for
execution later, it should preserve the thread context which is what
this change does.
Today we rewrite the operations from the leader with the term of the
following primary because the follower should own its history. The
problem is that a newly promoted primary may re-assign its term to
operations which were replicated to replicas before by the previous
primary. If this happens, some operations with the same seq_no may be
assigned different terms. This is not good for the future optimistic
locking using a combination of seqno and term.
This change ensures that the primary of a follower only processes an
operation if that operation was not processed before. The skipped
operations are guaranteed to be delivered to replicas via either
primary-replica resync or peer-recovery. However, the primary must not
acknowledge until the global checkpoint is at least the highest seqno of
all skipped ops (i.e., they all have been processed on every replica).
Relates #31751
Relates #31113
* New OCTET_LENGTH function
* Changed the way the FunctionRegistry stores functions, considering the alphabetic ordering by name
* Added documentation for the RANDOM function
This fixes an issue where an incorrect expected next step is used when checking
to execute `AsyncActionStep`s after a cluster state step.
It fixes this scenario:
- `ExecuteStepsUpdateTask` executes a `ClusterStateWaitStep` or
`ClusterStateActionStep` successfully
- The next step is also a `ClusterStateWaitStep`, so it loops
- The `ClusterStateWaitStep` has a next stepkey (which gets set to the
`nextStepKey` in the code)
- The `ClusterStateWaitStep` fails the condition, meaning that it will have to
wait longer
- The `nextStepKey` is now incorrect though, because we did not advance the
index's step, and it's not `null` (which is another safe value if there is no
step after the `ClusterStateWaitStep`)
This fixes the problem by resetting the nextStepKey to null if the condition is
not met, since we are not going to advance the step metadata in this
case (thereby skipping the `maybeRunAsyncAction` invocation).
This commit also tightens up and enhances much of the ILM logging. A lot of
logging was missing the index name (making it hard to debug in the presence of
multiple indices) and a lot was using the wrong logging level (DEBUG is now
actually readable without being a wall of text).
Resolves#34297
Since all calls to `ESLoggerFactory` outside of the logging package were
deprecated, it seemed like it'd simplify things to migrate all of the
deprecated calls and declare `ESLoggerFactory` to be package private.
This does that.
Also fixed ShardFollowNodeTaskTests to not return ops when responseSize
is empty. Otherwise ops are returned when no ops are expected to be returned.
Co-authored-by: Jason Tedor <jason@tedor.me>
Unfollow should be allowed / disallowed on a per index level instead of
cluster level.
Also renamed `create_follow_index` index privilege to
`manage_follow_index` privilege and include unfollow and close APIs.
This commit modifies the follow stats API response structure to more
clearly highlight meaning of the higher level fields. In particular,
previously the response had a top-level key for each index. Instead, we
nest the indices under an "indices" field which is now an array. The
values in this array are objects containing two fields: "index" which is
the name of the follower index, and "shards" which is an array where
each value in the array is the follower stats for that shard. That is,
we have gone from:
{
"bar": [
{
"shard_id": 0...
}...
]...
}
to
{
"indices": [
{
"index": "bar",
"shards": [
{
"shard_id": 0...
}...
]
}...
}
In the CCR docs we want to refer to the endpoint that returns following
stats as the follow stats API. This commit renames the internal
implementation of this endpoint to reflect this usage.
The "lookupUser" method on a realm facilitates the "run-as" and
"authorization_realms" features.
This commit allows a realm to be used for "lookup only", in which
case the "authenticate" method (and associated token methods) are
disabled.
It does this through the introduction of a new
"authentication.enabled" setting, which defaults to true.
Building automatons can be costly. For the most part we cache things
that use automatons so the cost is limited.
However:
- We don't (currently) do that everywhere (e.g. we don't cache role
mappings)
- It is sometimes necessary to clear some of those caches which can
cause significant CPU overhead and processing delays.
This commit introduces a new cache in the Automatons class to avoid
unnecesarily recomputing automatons.
There may be values in the thread context that ought to be preseved
for later use, even if one or more realms perform asynchronous
authentication.
This commit changes the AuthenticationService to wrap the potentially
asynchronous calls in a ContextPreservingActionListener that retains
the original thread context for the authentication.
This changes the delete job API by adding
the choice to delete a job asynchronously.
The commit adds a `wait_for_completion` parameter
to the delete job request. When set to `false`,
the action returns immediately and the response
contains the task id.
This also changes the handling of subsequent
delete requests for a job that is already being
deleted. It now uses the task framework to check
if the job is being deleted instead of the cluster
state. This is a beneficial for it is going to also
be working once the job configs are moved out of the
cluster state and into an index. Also, force delete
requests that are waiting for the job to be deleted
will not proceed with the deletion if the first task
fails. This will prevent overloading the cluster. Instead,
the failure is communicated better via notifications
so that the user may retry.
Finally, this makes the `deleting` property of the job
visible (also it was renamed from `deleted`). This allows
a client to render a deleting job differently.
Closes#32836
Drops the last logging constructor that takes `Settings` because it is
no longer needed.
Watcher goes through a lot of effort to pass `Settings` to `Logger`
constructors and dropping `Settings` from all of those calls allowed us
to remove quite a bit of log-based ceremony from watcher.
The Security plugin authorizes actions on indices. Authorization
happens on a per index/alias basis. Therefore a request with a
Multi Index Expression (containing wildcards) has to be
first evaluated in the authorization layer, before the request is
handled. For authorization purposes, wildcards in expressions will
only be expanded to indices/aliases that are visible by the authenticated
user. However, this "constrained" evaluation has to be compatible with
the expression evaluation that a cluster without the Security plugin
would do. Therefore any change in the evaluation logic
in any of these sites has to be mirrored in the other site.
This commit mirrors the changes in core from #33518 that allowed
for Multi Index Expression in the Get Alias API, loosely speaking.
This enables Elasticsearch to use the JVM-wide configured
PKCS#11 token as a keystore or a truststore for its TLS configuration.
The JVM is assumed to be configured accordingly with the appropriate
Security Provider implementation that supports PKCS#11 tokens.
For the PKCS#11 token to be used as a keystore or a truststore for an
SSLConfiguration, the .keystore.type or .truststore.type must be
explicitly set to pkcs11 in the configuration.
The fact that the PKCS#11 token configuration is JVM wide implies that
there is only one available keystore and truststore that can be used by TLS
configurations in Elasticsearch.
The PIN for the PKCS#11 token can be set as a truststore parameter in
Elasticsearch or as a JVM parameter ( -Djavax.net.ssl.trustStorePassword).
The basic goal of enabling PKCS#11 token support is to allow PKCS#11-NSS in
FIPS mode to be used as a FIPS 140-2 enabled Security Provider.
When the cluster.routing.allocation.disk.watermark.flood_stage watermark
is breached, DiskThresholdMonitor marks the indices as read-only. This
failed when x-pack security was present as system user does not have the privilege
for update settings action("indices:admin/settings/update").
This commit adds the required privilege for the system user. Also added missing
debug logs when access is denied to help future debugging.
An assert statement is added to catch any missed privileges required for
system user.
Closes#33119
In SessionFactoryLoadBalancingTests#testRoundRobinWithFailures()
we kill ldap servers randomly and immediately bind to that port
connecting to mock server socket. This is done to avoid someone else
listening to this port. As the creation of mock socket and binding to the
port is immediate, sometimes the earlier socket would be in TIME_WAIT state
thereby having problems with either bind or connect.
This commit sets the SO_REUSEADDR explicitly to true and also sets
the linger on time to 0(as we are not writing any data) so as to
allow re-use of the port and close immediately.
Note: I could not find other places where this might be problematic
but looking at test runs and netstat output I do see lot of sockets
in TIME_WAIT. If we find that this needs to be addressed we can
wrap ServerSocketFactory to set these options and use that with in
memory ldap server configuration during tests.
Closes#32190
This commit upgrades the unboundid ldapsdk to version 4.0.8. The
primary driver for upgrading is a fix that prevents this library from
rewrapping Error instances that would normally bubble up to the
UncaughtExceptionHandler and terminate the JVM. Other notable changes
include some fixes related to connection handling in the library's
connection pool implementation.
Closes#33175
- this adds an integration test that runs through a policy
with all the actions defined.
- adds a test specific to a policy having just a rollover action
- bumps the node count to 4
The `DnRoleMapper` class is used to map distinguished names of groups
and users to role names. This mapper builds in an internal map that
maps from a `com.unboundid.ldap.sdk.DN` to a `Set<String>`. In cases
where a lot of distinct DNs are mapped to roles, this can consume quite
a bit of memory. The majority of the memory is consumed by the DN
object. For example, a 94 character DN that has 9 relative DNs (RDN)
will retain 4KB of memory, whereas the String itself consumes less than
250 bytes.
In order to reduce memory usage, we can map from a normalized DN string
to a List of roles. The normalized string is actually how the DN class
determines equality with another DN and we can drop the overhead of
needing to keep all of the other objects in memory. Additionally the
use of a List provides memory savings as each HashSet is backed by a
HashMap, which consumes a great deal more memory than an appropriately
sized ArrayList. The uniqueness we get from a Set is maintained by
first building a set when parsing the file and then converting to a
list upon completion.
Closes#34237
Today we reverse the initial order of the nested documents when we
index them in order to ensure that parents documents appear after
their children. This means that a query will always match nested documents
in the reverse order of their offsets in the source document.
Reversing all documents is not needed so this change ensures that parents
documents appear after their children without modifying the initial order
in each nested level. This allows to match children in the order of their
appearance in the source document which is a requirement to efficiently
implement #33587. Old indices created before this change will continue
to reverse the order of nested documents to ensure backwark compatibility.
This commit changes the way that step execution flows. Rather than have any step
run when the cluster state changes or the periodic scheduler fires, this now
runs the different types of steps at different times.
`AsyncWaitStep` is run at a periodic manner, ie, every 10 minutes by default
`ClusterStateActionStep` and `ClusterStateWaitStep` are run every time the
cluster state changes.
`AsyncActionStep` is now run only after the cluster state has been transitioned
into a new step. This prevents these non-idempotent steps from running at the
same time. It addition to being run when transitioned into, this is also run
when a node is newly elected master (only if set as the current step) so that
master failover does not fail to run the step.
This also changes the `RolloverStep` from an `AsyncActionStep` to an
`AsyncWaitStep` so that it can run periodically.
Relates to #29823
The follower index shard history UUID will be fetched from the indices stats api when the shard follow task starts and will be provided with the bulk shard operation requests. The bulk shard operations api will fail if the provided history uuid is unequal to the actual history uuid.
No longer record the leader history uuid in shard follow task params, but rather use the leader history UUIDs directly from follower index's custom metadata. The resume follow api will remain to fail if leader index shard history UUIDs are missing.
Closes#33956
Revert "[TESTS] Pin MockWebServer to TLS1.2 (#33127)" (commit
214652d4af) and "Pin TLS1.2 in
SSLConfigurationReloaderTests" (commit
d9f5e4fd2e), which pinned the
MockWebServer used in the SSLConfigurationReloaderTests to TLSv1.2 in
order to prevent failures with JDK 11 related to ssl session
invalidation. We no longer need this pinning as the problematic code
was fixed in #34130.
The `-` and `+` as a number literal prefix are already
parsed by the rule in `valueExpression`. To accommodate
this, there are some code changes that enables the
`ExpressionBuilder` to parse Literal integers and decimals
together with the `-/+` prefix sign (if exists) and validate
them (wrong format, large numbers, etc.).
Follows: #33854
Adds support for the get rollup job to the High Level REST Client. I had
to do three interesting and unexpected things:
1. I ported the rollup state wiping code into the high level client
tests. I'll move this into the test framework in a followup and remove
the x-pack version.
2. The `timeout` in the rollup config was serialized using the
`toString` representation of `TimeValue` which produces fractional time
values which are more human readable but aren't supported by parsing. So
I switched it to `getStringRep`.
3. Refactor the xcontent round trip testing utilities so we can test
parsing of classes that don't implements `ToXContent`.
Previously, parsing an arithmetic expression with `*` and no spaces,
e.g.: `2*i` threw a parsing exception as the grammar rule for
tableIdentifier was clashing with the rule for arithmetic operator `*`.
This issue comes already in the lexer and the left part of the
expression (in our example `2*`) was recognised as a
TABLE_IDENTIFIER token.
The solution adopted is to allow the `*` wildcard in the table name
only if it's surrounded with double quotes, e.g.: `"my*index"`
Closes: #33957
Remove CamelCase to CAMEL_CASE conversion when resolving
a function. Only convert user input to upper case and then
try to match with aliases or primary names.
Keep the internal conversion FunctionName to FUNCTION__NAME
which provides flexibility when registering functions by their class
name.
Fixes: #34114
Since #34099, the FollowingEngine will skip an operation which was
already processed before. With that change, it should be okay to unmute
testFollowIndexAndCloseNode.
This change fixes a potential deadlock problem in the unit
test introduced in #34117.
It also removes a piece of debug code and corrects a docs
formatting problem that were both added in that same PR.
This arose when two commits were pushed at roughly the same time, both
of which compiled successfully against master, but not when taken
together. This commit fixes a reference in one of the commits that was
changed in the other commit.
This commit modifies the CCR stats endpoint for indices to be
/{index}/_ccr/stats. This makes this endpoint consistent with other
index-centric endpoints like indices stats.
The unfollow API changes a follower index into a regular index, so that it will accept write requests from clients.
For the unfollow api to work the index follow needs to be stopped and the index needs to be closed.
Closes#33931
This change introduces the indexing optimization using sequence numbers
in the FollowingEngine. This optimization uses the max_seq_no_updates
which is tracked on the primary of the leader and replicated to replicas
and followers.
Relates #33656
This can be used to restrict the amount of CPU a single
structure finder request can use.
The timeout is not implemented precisely, so requests
may run for slightly longer than the timeout before
aborting.
The default is 25 seconds, which is a little below
Kibana's default timeout of 30 seconds for calls to
Elasticsearch APIs.
Prior to following an index in the follow API, check whether current
user has sufficient privileges in the leader cluster to read and
monitor the leader index.
Also check this in the create and follow API prior to creating the
follow index.
Also introduced READ_CCR cluster privilege that include the minimal
cluster level actions that are required for ccr in the leader cluster.
So a user can follow indices in a cluster, but not use the ccr admin APIs.
Closes#33553
Co-authored-by: Jason Tedor <jason@tedor.me>
In prior versions of Java, we expected to see a SSLHandshakeException
when starting a handshake with a server that we do not trust. In JDK11,
the exception has changed to a SSLException, which
SSLHandshakeException extends. This is most likely a side effect of the
TLS 1.3 changes in JDK11. This change updates the test to catch the
SSLException instead of the SSLHandshakeException and enables the test
to work on JDK8 through JDK11.
Closes#29989
The SSLService invalidates SSLSessions when there is a change to any of
the underlying key or trust material. However, this invalidation code
did not check for a null SSLSession being returned from the context and
assumed that the context would always return a non-null object. The
return of a null object is possible in all versions, but JDK11 seems to
return them more often due to changes for TLS 1.3. There are a number
of reasons that we get a id of a session but the context returns null
when the session with that id is requested. Some of the reasons for
this are:
* Session was evicted by session cache
* Session has timed out
* Session has been invalidated by another caller
To handle this, the SSLService now checks if the value is null before
calling invalidate on the SSLSession.
Closes#32124
This commit removes the use of ExecutableScript from watcher in favor of
custom script contexts for both watcher condition scripts and transform
scripts.
In prior versions of Java, we expected to see a SSLHandshakeException
when starting a handshake with a server that we do not trust. In JDK11,
the exception has changed to a SSLException, which
SSLHandshakeException extends. This is most likely a side effect of the
TLS 1.3 changes in JDK11. This change updates the test to catch the
SSLException instead of the SSLHandshakeException and enables the test
to work on JDK8 through JDK11.
Closes#32293
The following changes were made:
* Added ElasticsearchSecurityException. For in the case the current user has insufficient privileges while an index is being followed. Prior to following ccr checks whether the current user has sufficient privileges and if not the follow api fails with an error.
* Added Index block exception. If the leader index gets closed, this exception is returned.
* Added ClusterBlockException service unavailable. In case for example the leader cluster is without elected master.
* Removed IndexNotFoundException. If the leader / follower index has been deleted, ccr will need to stop the shard follow tasks with an error.
Closes#33954
Fixes the equals and hash function to ignore the order of aggregations to ensure equality after serialization
and deserialization. This ensures storing configs with aggregation works properly.
This also addresses a potential issue in caching when the same query contains aggregations but in
different order. 1st it will not hit in the cache, 2nd cache objects which shall be equal might end up twice in
the cache.
Due to a bug, that was fixed in #33360 and commit
1de2a925ce the initial adding of a watch
could get lost, thus leaving the watcher stats count as zero despite adding a watch.
Closes#33326
* Renamed CCR APIs
Renamed:
* `/{index}/_ccr/create_and_follow` to `/{index}/_ccr/follow`
* `/{index}/_ccr/unfollow` to `/{index}/_ccr/pause_follow`
* `/{index}/_ccr/follow` to `/{index}/_ccr/resume_follow`
Relates to #33931
always use `IndicesOptions.strictExpand()` for indices options.
The follow index may be closed and we still want to get stats from
shard follow task and the whether the provided index name matches with
follow index name is checked when locating the task itself in the ccr
stats transport action.
`Settings` is no longer required to get a `Logger` and we went to quite
a bit of effort to pass it to the `Logger` getters. This removes the
`Settings` from all of the logger fetches in security and x-pack:core.
Centralize and simplify the script generation between operators and functions which are currently decoupled. As part of this process most predicates (<, <=, etc...) were made ScalarFunction as their purpose and functionality is quite similar (see % and MOD functions).
Renamed ProcessDefinition to Pipe
Add ScriptWeaver as a mixin-in interface for script customization
Add logic for equals/lte/lt
Improve BinaryOperator/expression toString
Minimize duplication across string functions
Close#33975
This commit adds the ability to plug in compilation of custom contexts
in mock script engine. This is needed for testing plugins which add
custom contexts like watcher.
Watcher is using a lot of so called TextTemplate fields in a watch
definition, which can use mustache to insert the watch id for example.
For the user it is non-obvious which field is just a string field or
which field is a text template.
This also means, that for every such field, we currently do a script
compilation, even if the field does not contain any mustache syntax.
This will lead to an increased script cache churn, because those
compiled scripts (that only contain a string), will evict other scripts.
On top of that, this also means that an unneeded compilation has
happened, instead of returning that string immediately.
The usages of mustache templating are in all of the actions (most of the time far
more than one compilation) as well as most of the inputs.
Especially when running a lot of watches in parallel, this will reduce
execution times and help reuse of real scripts.
Security previously hardcoded a default scroll keepalive of 10 seconds,
but in some cases this is not enough time as there can be network
issues or overloading of host machines. After this change, security
will now use the default keepalive timeout, which is controllable using
a setting and the default value is 5 minutes.
In order to optimize the use of the role cache, when the roles.yml file
is reloaded we now calculate the names of removed, changed, and added
roles so that they may be passed to any listeners. This allows a
listener to selectively clear cache for only the roles that have been
modified. The CompositeRolesStore has been adapted to do exactly that
so that we limit the need to reload roles from sources such as the
native roles stores or external role providers.
See #33205
This commits creates a DateMathParser interface, which is already
implemented for both joda and java time. While currently the java time
DateMathParser is not used, this change will allow a followup which will
create a DateMathParser from a DateFormatter, so the caller does not
need to know the internals of the DateFormatter they have.
This change cleans up "unused variable" warnings. There are several cases were we
most likely want to suppress the warnings (especially in the client documentation test
where the snippets contain many unused variables). In a lot of cases the unused
variables can just be deleted though.
This commit replicates the max_seq_no_of_updates on the leading index
to the primaries of the following index via ShardFollowNodeTask. The
max_seq_of_updates is then transmitted to the replicas of the follower
via replication requests (that's BulkShardOperationsRequest).
Relates #33656
Implement circuit breaker logic in the parser which catches expressions
that can blow up the tree and result in StackOverflowError being thrown.
Co-authored-by: Costin Leau <costin.leau@gmail.com>
ILM now only forces indices to become read only in the case of Shrink
and Force Merge actions, as these are most useful in cases where the
index is no longer being written to.
[ML] Modify thresholds for normalization triggers
The (arbitrary) threshold factors used to judge if scores have
changed significantly enough to trigger a look-back renormalization have
been changed to values that reduce the frequency of such
renormalizations.
Added a clause to treat changes in scores as a 'big change' if it would
result in a change of severity reported in the UI.
Also altered the clause affecting small scores so that a change should
be considered big if scores have changed by at least 1.5.
Relates https://github.com/elastic/machine-learning-qa/issues/263
We start tracking max seq_no_of_updates on the primary in #33842. This
commit replicates that value from a primary to its replicas in replication
requests or the translog phase of peer-recovery.
With this change, we guarantee that the value of max seq_no_of_updates
on a replica when any index/delete operation is performed at least the
max_seq_no_of_updates on the primary when that operation was executed.
Relates #33656
Previously the timestamp_formats field in the response
from the find_file_structure endpoint contained Joda
timestamp formats. This change makes that clear by
renaming the field to joda_timestamp_formats, and also
adds a java_timestamp_formats field containing the
equivalent Java time format strings.
Two issues are resolved:
1. The `value_type` should be either long or double in case of numeric.
2. The key label for the aggregate filter (having) was duplicate of an aggr key label.
Fixes: #33520
* TESTS: Stabilize Renegotiation Test
* The second `startHandshake` is not synchronous and a read of only
50ms may fail to trigger it entirely (the failure can be reproduced reliably by setting the socket timeout to `1`)
=> fixed by retrying the read until the handshake finishes (a longer timeout would've worked too,
but retrying seemed more stable)
* Closes#33772
* Make certain ML node settings dynamic (#33565)
* Changing to pull in updating settings and pass to constructor
* adding note about only newly opened jobs getting updated value
This commit introduces an AbstractSimpleSecurityTransportTestCase for
security transports. This classes provides transport tests that are
specific for security transports. Additionally, it fixes the tests referenced in
#33285.
The source only snapshot drops fully deleted segments before snapshotting
them. In order to compare them we need to drop all fully deleted segments
in the test as well.
Closes#33755
If numWrites is between 2 and 9, we will issue an invalid range because
the from_seq_no is negative. This commit makes sure that numWrites is at
least 10, and adds an explicit test to verify invalid request ranges.
SearchGroupsResolverInMemoryTests was (rarely) fail in a way that
suggests that the server-side delay (100ms) was not enough to trigger
the client-side timeout (5ms).
The server side delay has been increased to try and overcome this.
Resolves: #32913
Previously numeric values in the field_stats created by the
find_file_structure endpoint were always output with a
decimal point. This looked unfriendly and unnatural for
fields that clearly store integer values. This change
converts integer values to type Integer before output in
the file structure field stats.
This PR is the first step to use seq_no to optimize indexing operations.
The idea is to track the max seq_no of either update or delete ops on a
primary, and transfer this information to replicas, and replicas use it
to optimize indexing plan for index operations (with assigned seq_no).
The max_seq_no_of_updates on primary is initialized once when a primary
finishes its local recovery or peer recovery in relocation or being
promoted. After that, the max_seq_no_of_updates is only advanced internally
inside an engine when processing update or delete operations.
Relates #33656
This commit reverts most of #33157 as it introduces another race
condition and breaks a common case of watcher, when the first watch is
added to the system and the index does not exist yet.
This means, that the index will be created, which triggers a reload, but
during this time the put watch operation that triggered this is not yet
indexed, so that both processes finish roughly add the same time and
should not overwrite each other but act complementary.
This commit reverts the logic of cleaning out the ticker engine watches
on start up, as this is done already when the execution is paused -
which also gets paused on the cluster state listener again, as we can be
sure here, that the watches index has not yet been created.
This also adds a new test, that starts a one node cluster and emulates
the case of a non existing watches index and a watch being added, which
should result in proper execution.
Closes#33320
This change adds the OneStatementPerLineCheck to our checkstyle precommit
checks. This rule restricts the number of statements per line to one. The
resoning behind this is that it is very difficult to read multiple statements on
one line. People seem to mostly use it in short lambdas and switch statements in
our code base, but just going through the changes already uncovered some actual
problems in randomization in test code, so I think its worth it.
The job deletion logic was scattered around a few places:
the transport action, the job manager and the deletion task.
Overloading the task with deletion logic also meant extra
dependencies in the core package which should be unnecessary.
This commit consolidates all this logic into the transport action
and replaces the deletion task with a plain one that needs not be
aware of deletion logic.
* Added TRUNCATE function, modified ROUND to accept two parameters instead of one. Made the second parameter optional for both functions.
* Added documentation for both functions.
Removed rules in the grammar that were superfluous, as they
are already "caught" other rules in the same context.
Also switched to exact ambig detection for debug mode
Fixes: #31885
Using index settings for ILM state is fragile and exposes too much
information that doesn't need to be exposed. Using custom index metadata
is more resilient and allows more controlled access to internal
information.
As part of these changes, moves away from using defaults for ILM-related
values, in favor of using null values to clearly indicate that the value is not
present.
Changes the default of the `node.name` setting to the hostname of the
machine on which Elasticsearch is running. Previously it was the first 8
characters of the node id. This had the advantage of producing a unique
name even when the node name isn't configured but the disadvantage of
being unrecognizable and not being available until fairly late in the
startup process. Of particular interest is that it isn't available until
after logging is configured. This forces us to use a volatile read
whenever we add the node name to the log.
Using the hostname is available immediately on startup and is generally
recognizable but has the disadvantage of not being unique when run on
machines that don't set their hostname or when multiple elasticsearch
processes are run on the same host. I believe that, taken together, it
is better to default to the hostname.
1. Running multiple copies of Elasticsearch on the same node is a fairly
advanced feature. We do it all the as part of the elasticsearch build
for testing but we make sure to set the node name then.
2. That the node.name defaults to some flavor of "localhost" on an
unconfigured box feels like it isn't going to come up too much in
production. I expect most production deployments to at least set the
hostname.
As a bonus, production deployments need no longer set the node name in
most cases. At least in my experience most folks set it to the hostname
anyway.
We currently special-case SynonymFilterFactory and SynonymGraphFilterFactory, which need to
know their predecessors in the analysis chain in order to correctly analyze their synonym lists. This
special-casing doesn't work with Referring filter factories, such as the Multiplexer or Conditional
filters. We also have a number of filters (eg the Multiplexer) that will break synonyms when they
appear before them in a chain, because they produce multiple tokens at the same position.
This commit adds two methods to the TokenFilterFactory interface.
* `getChainAwareTokenFilterFactory()` allows a filter factory to rewrite itself against its preceding
filter chain, or to resolve references to other filters. It replaces `ReferringFilterFactory` and
`CustomAnalyzerProvider.checkAndApplySynonymFilter`, and by default returns `this`.
* `getSynonymFilter()` defines whether or not a filter should be applied when building a synonym
list `Analyzer`. By default it returns `true`.
Fixes#33609
The fix in #33757 introduces some workaround since FilterCodecReader didn't
support unwrapping. This cuts over to a more elegant fix to access the readers
segment infos.
Previously multiple comma separated lists of options where not
recognized correctly which resulted in only the last of them
to be taked into account, e.g.:
For the following query:
SELECT * FROM test WHERE QUERY('search', 'default_field=foo', 'default_operator=and')"
only the `default_operator=and` was finally passed to the ES query.
Fixes: #32602
Instead of having one constructor that accepts all arguments, all parameters
should be provided via setters. Only leader and follower index are required
arguments. This makes using this class in tests and transport client easier.
DAYNAME and MONTHNAME functions tests will be skipped if the right JVM parameter (-Djava.locale.providers=COMPAT) is not used in unit testing environment
This moves away from caching a list of steps for a current phase, instead
rebuilding the necessary step from the phase JSON stored in the index's
metadata.
Relates to #29823
Currently a watch execution results in one bulk request, when the
triggered watches are written into the that index, that need to be
executed. However the update of the watch status, the creation of the
watch history entry as well as the deletion of the triggered watches
index are all single document operations.
This can have quite a negative impact, once you are executing a lot of
watches, as each execution results in 4 documents writes, three of them
being single document actions.
This commit switches to a bulk processor instead of a single document
action for writing watch history entries and deleting triggered watch
entries. However the defaults are to run synchronous as before because
the number of concurrent requests is set to 0. This also fixes a bug,
where the deletion of the triggered watch entry was done asynchronously.
However if you have a high number of watches being executed, you can
configure watcher to delete the triggered watches entries as well as
writing the watch history entries via bulk requests.
The triggered watches deletions should still happen in a timely manner,
where as the history entries might actually be bound by size as one
entry can easily have 20kb.
The following settings have been added:
- xpack.watcher.bulk.actions (default 1)
- xpack.watcher.bulk.concurrent_requests (default 0)
- xpack.watcher.bulk.flush_interval (default 1s)
- xpack.watcher.bulk.size (default 1mb)
The drawback of this is of course, that on a node outage you might end
up with watch history entries not being written or watches needing to be
executing again because they have not been deleted from the triggered
watches index. The window of these two cases increases configuring the bulk processor to wait to reach certain thresholds.
The following stats are being kept track of:
1) The total number of times that auto following a leader index succeed.
2) The total number of times that auto following a leader index failed.
3) The total number of times that fetching a remote cluster state failed.
4) The most recent 256 auto follow failures per auto leader index
(e.g. create_and_follow api call fails) or cluster alias
(e.g. fetching remote cluster state fails).
Each auto follow run now produces a result that is being used to update
the stats being kept track of in AutoFollowCoordinator.
Relates to #33007
* Implement xpack.monitoring.elasticsearch.collection.enabled setting
* Fixing line lengths
* Updating constructor calls in test
* Removing unused import
* Fixing line lengths in test classes
* Make monitoringService.isElasticsearchCollectionEnabled() return true for tests
* Remove wrong expectation
* Adding unit tests for new flag to be false
* Fixing line wrapping/indentation for better readability
* Adding docs
* Fixing logic in ClusterStatsCollector::shouldCollect
* Rebasing with master and resolving conflicts
* Simplifying implementation by gating scheduling
* Doc fixes / improvements
* Making methods package private
* Fixing wording
* Fixing method access
This PR removes fields that are not actually used by the Monitoring UI. This will greatly simplify the eventual migration to using Metricbeat for monitoring Elasticsearch (see https://github.com/elastic/beats/pull/8260#discussion_r215885868 for more context and discussion around removing these fields from ES collection).
* [CCR] Do not unnecessarily wrap fetch exception in a ElasticSearch exception and
properly map fetch_exception.exception field as object.
The extra caused by level is not necessary here:
```
"fetch_exceptions": [
{
"from_seq_no": 1,
"retries": 106,
"exception": {
"type": "exception",
"reason": "[index1] IndexNotFoundException[no such index]",
"caused_by": {
"type": "index_not_found_exception",
"reason": "no such index",
"index_uuid": "_na_",
"index": "index1"
}
}
}
],
```
When a leader index is created, it may not have a mapping yet.
Currently if you follow such an index the shard follow tasks fail with
NoSuchElementException, because they expect a single mapping.
This commit fixes that, by allowing that a leader index does not yet have
a mapping.
We can't rely on the leaf reader ordinal in a wrapped reader since
it might not correspond to the ordinal in the SegmentInfos for it's
SegmentCommitInfo.
Relates to #32844Closes#33689Closes#33755
This commit adds the Create Rollup Job API to the high level REST
client. It supersedes #32703 and adds dedicated request/response
objects so that it does not depend on server side components.
Related #29827
Rather than scheduling pings to the leader index when we are caught up
to the leader, this commit introduces long polling for changes. We will
fire off a request to the leader which if we are already caught up will
enter a poll on the leader side to listen for global checkpoint
changes. These polls will timeout after a default of one minute, but can
also be specified when creating the following task. We use these time
outs as a way to keep statistics up to date, to not exaggerate time
since last fetches, and to avoid pipes being broken.
When executing CCR REST tests it is going to be expected after global
checkpoint polling goes in that shard changes tasks can still be pending
at the end of the test. One way to deal with this is to set a low
timeout on these polls, but then that means we are not executing our
REST tests with our default production settings and instead would be
using an unrealistic low timeout. Alternatively, since we expect these
tasks to be there, we can not count them against the test. That is what
this commit does.
This commit moves these REST tests (possibly temporarily) to a
sub-project of ccr. We do this (again, possibly temporarily) to keep
them within the ccr sub-project yet there are changes within 6.x that
prevent these from being in the top-level project (the cluster formation
tasks are trying to install x-pack-ccr into the
integ-test-zip). Therefore, we isolate these for now until we can
understand why there are differences between 6.x and master.
There was a listener that re-runs the policy with the new state when the cluster
state is processed by the `MoveToNextStepUpdateTask`. This removes this listener
as we will execute the policy through the `IndexLifecyleService` cluster state
listener.
When developing ccr it is not ideal if tests are in multiple modules.
Even the classes these tests test are in the core module, it is easier
if these tests are in ccr module in order to avoid running the test task
in core module. This results in running many non ccr tests.
This way when developing ccr we can run locally:
./gradlew x-pack:plugin:core:precommit x-pack:plugin:ccr:check
before pushing to PR branches and be confident that the PR build passes,
without running x-pack:plugin:core:check task.
Ensure that the SSLConfigurationReloaderTests can run with JDK 11
by pinning the HttpClient to TLS version to TLS1.2. This is necessary
becase even if the MockWebServer is set to user TLS1.2, we don't
set its enabled protocols, so if it receives a TLS1.3 request (which
is the default behavior for HttpClient in JDK11), it will use TLS1.3
and the original issue will manifest again.
Relates #33127Resolves#32124
Changes the format of log events in the audit logfile.
It also changes the filename suffix from `_access` to `_audit`.
The new entry format is consistent with Elastic Common Schema.
Entries are formatted as JSON with no nested objects and field
names have a dotted syntax. Moreover, log entries themselves
are not spaced by commas and there is exactly one entry per line.
In addition, entry fields are ordered, unlike a typical JSON doc,
such that a human would not strain his eyes over jumbled
fields from one line to the other; the order is defined in the log4j2
properties file.
The implementation utilizes the log4j2's `StringMapMessage`.
This means that the application builds the log event as a map
and the log4j logic (the appender's layout) handle the format
internally. The layout, such as the set of printed fields and their
order, can be changed at runtime without restarting the node.
and if so debug log it and otherwise rethrow.
This should fix a couple of test failures where during test teardown tests
failed due to uncaught exceptions being detected.
This change modifies the file structure detection functionality
such that some of the decisions can be overridden with user
supplied values.
The fields that can be overridden are:
- charset
- format
- has_header_row
- column_names
- delimiter
- quote
- should_trim_fields
- grok_pattern
- timestamp_field
- timestamp_format
If an override makes finding the file structure impossible then
the endpoint will return an exception.
We have a Kerberos setting to remove realm part from the user
principal name (remove_realm_name). If this is true then
the realm name is removed to form username but in the process,
the realm name is lost. For scenarios like Kerberos cross-realm
authentication, one could make use of the realm name to determine
role mapping for users coming from different realms.
This commit adds user metadata for kerberos_realm and
kerberos_user_principal_name.
Disable specific Thai and Japanese locales as Certificate expiration
validation fails due to the date parsing of BouncyCastle (that manifests
in a FIPS 140 JVM as this is the only place we use BouncyCastle).
Added the locale switching logic here instead of subclassing
ESTestCase as these are the only tests that fail for these locales and
JVM combination.
Resolves#33081
We have a test dependency on Apache Mina when using SimpleKdcServer
for testing Kerberos. When checking for LDAP backend connectivity,
the code checks for deadlocks which require additional security
permissions accessClassInPackage.sun.reflect. As this is only for
test and we do not want to add security permissions to production,
this commit moves these tests and related classes to
x-pack evil-tests where they can run with security manager disabled.
The plan is to handle the security manager exception in the upstream issue
DIRMINA-1093
and then once the release is available to run these tests with security
manager enabled.
Closes#32739
This change removes the wrapping of the created field in the put user
response. The created field was added as a top level field in #32332,
while also still being wrapped within the `user` object of the
response. Since the value is available in both formats in 6.x, we can
remove the wrapped version for 7.0.
The follow index api checks if the recorded uuid in the follow index matches
with uuid of the leader index and fails otherwise. This validation will
prevent a follow index from following an incompatible leader index.
The create_and_follow api will automatically add this custom index metadata
when it creates the follow index.
Closes#31505
Previously, when an non-pruned cast (casting as a different
data type) got applied on a table column in the `SELECT` clause,
the name of the result column didn't contain the target data type
of the cast, e.g.:
SELECT CAST(MAX(salary) AS DOUBLE) FROM "test_emp"
returned as column name:
CAST(MAX(salary))
instead of:
CAST(MAX(salary) AS DOUBLE)
Closes#33571
* Added more tests for trivial casts that are pruned
Today we use a special unicast hosts provider, the `MockUncasedHostsProvider`,
in many integration tests, to deal with the dynamic nature of the allocation of
ports to nodes. However #33241 allows us to use file-based discovery to achieve
the same goal, so the special test-only `MockUncasedHostsProvider` is no longer
required.
This change removes `MockUncasedHostProvider` and replaces it with file-based
discovery in tests based on `EsIntegTestCase`.
Follow up to #33617. Relates to #30086.
As with all other per-index Monitoring collectors, the `CcrStatsCollector` should only collect stats for the indices the user wants to monitor. This list is controlled by the `xpack.monitoring.collection.indices` setting and defaults to all indices.
This change addresses some issues regarding thread safety around
updates and method calls on the XPackLicenseState object. There exists
a possibility that there could be a concurrent update to the
XPackLicenseState when there is a scheduled check to see if the license
is expired and a cluster state update. In order to address this, the
update method now has a synchronized block where member variables are
updated. Each method that reads these variables is now also
synchronized.
Along with the above change, there was a consistency issue around
security calls to the license state. The majority of security checks
make two calls to the license state, which could result in incorrect
behavior due to the checks being made against different license states.
The majority of this behavior was introduced for 6.3 with the inclusion
of x-pack in the default distribution. In order to resolve the majority
of these cases, the `isSecurityEnabled` method is no longer public and
the logic is also included in individual methods about security such as
`isAuthAllowed`. There were a few cases where this did not remove
multiple calls on the license state, so a new method has been added
which creates a copy of the current license state that will not change.
Callers can use this copy of the license state to make decisions based
on a consistent view of the license state.
For correctness we need to verify whether the history uuid of the leader
index shards never changes while that index is being followed.
* The history UUIDs are recorded as custom index metadata in the follow index.
* The follow api validates whether the current history UUIDs of the leader
index shards are the same as the recorded history UUIDs.
If not the follow api fails.
* While a follow index is following a leader index; shard follow tasks
on each shard changes api call verify whether their current history uuid
is the same as the recorded history uuid.
Relates to #30086
Co-authored-by: Nhat Nguyen <nhat.nguyen@elastic.co>
This change adds a `_source` only snapshot repository that allows to wrap
any existing repository as a _backend_ to snapshot only the `_source` part
including live docs markers. Snapshots taken with the `source` repository
won't include any indices, doc-values or points. The snapshot will be reduced in size and
functionality such that it requires full re-indexing after it's successfully restored.
The restore process will copy the `_source` data locally starts a special shard and engine
to allow `match_all` scrolls and searches. Any other query, or get call will fail with and unsupported operation exception. The restored index is also marked as read-only.
This feature aims mainly for disaster recovery use-cases where snapshot size is
a concern or where time to restore is less of an issue.
**NOTE**: The snapshot produced by this repository is still a valid lucene index. This change doesn't allow for any longer retention policies which is out of scope for this change.
Improve failure handling of retryable errors by retrying remote calls in
a exponential backoff like manner. The delay between a retry would not be
longer than the configured max retry delay. Also retryable errors will be
retried indefinitely.
Relates to #30086
* SQL: Make Literal a NamedExpression
Literal now is a NamedExpression reducing the need for Aliases for
folded expressions leading to simpler optimization rules.
Fix#33523
This change tightens up the meaning of the "input_fields" field
in the file structure finder output. Previously it was permitted
but not calculated for JSON and XML files. Following this change
the field is called "column_names" and is only permitted for
delimited files.
Additionally the way the column names are set for headerless
delimited files is refactored to encapsulate the way they're
named to one line of the code rather than having the same
logic in two places.
Previously, when an arithmetic function got applied on a
table column in the `SELECT` clause, the name of the result
column contained weird characters used internally when
processing the SQL statement e.g.:
SELECT CHAR(emp_no % 10000) FROM "test_emp"
returned:
CHAR((emp_no{f}#14) % 10000))
as the column name instead of:
CHAR((emp_no) % 10000))
Also, fix an issue that causes a ClassCastException to be thrown
when using functions where both arguments are literals.
Closes#31869Closes#33461
* LeafCollector.setScorer() now takes a Scorable
* Scorers may not have null Weights
* IndexWriter.getFlushingBytes() reports how much memory is being used by IW threads writing to disk
* [CCR] Delay auto follow license check
so that we're sure that there are auto follow patterns configured
Otherwise we log a warning in case someone is running with basic or gold
license and has not used the ccr feature.
This is a new index privilege that the user needs to have in the follow cluster.
This privilege is required in addition to the `manage_ccr` cluster privilege in
order to execute the create and follow api.
Closes#33555
* Correctly handle NONE keyword for system keystore
As defined in the PKCS#11 reference guide
https://docs.oracle.com/javase/8/docs/technotes/guides/security/p11guide.html
PKCS#11 tokens can be used as the JSSE keystore and truststore and
the way to indicate this is to set `javax.net.ssl.keyStore` and
`javax.net.ssl.trustStore` to `NONE` (case sensitive).
This commits ensures that we honor this convention and do not
attempt to load the keystore or truststore if the system property is
set to NONE.
* Handle password protected system truststore
When a PKCS#11 token is used as the system truststore, we need to
pass a password when loading it, even if only for reading
certificate entries. This commit ensures that if
`javax.net.ssl.trustStoreType` is set to `PKCS#11` (as it would
when a PKCS#11 token is in use) the password specified in
`javax.net.ssl.trustStorePassword` is passed when attempting to
load the truststore.
Relates #33459
In some cases we want to deprecate a setting, and then automatically
upgrade uses of that setting to a replacement setting. This commit adds
infrastructure for this so that we can upgrade settings when recovering
the cluster state, as well as when such settings are dynamically applied
on cluster update settings requests. This commit only focuses on cluster
settings, index settings can build on this infrastructure in a
follow-up.
When requesting job stats for `_all`, all ES tasks are accepted
resulting to loads of cluster traffic and a memory overhead.
This commit correctly filters out non ML job tasks.
Closes#33515
We may use different global checkpoints to validate/normalize the range
of a change request if the global checkpoint is advanced between these
calls. If this is the case, then we generate an invalid request range.
This commit reverses the logic for CCR license checks in a few
actions. This is done so that the successful case, which tends to be a
larger block of code, does not require indentation.
We have some listeners in the CCR license tests that invoke Assert#fail
if the onSuccess method for the listener is unexpectedly invoked. This
can leave the main test thread hanging until the test suite times out
rather than failing quickly. This commit adds some latch countdowns so
that we fail quickly if these cases are hit.
In the multi-cluster-with-non-compliant-license tests, we try to write
out a java.policy to a temporary directory. However, if this temporary
directory does not already exist then writing the java.policy file will
fail. This commit ensures that the temporary directory exists before we
attempt to write the java.policy file.
This commit adds license checks for the auto-follow implementation. We
check the license on put auto-follow patterns, and then for every
coordination round we check that the local and remote clusters are
licensed for CCR. In the case of non-compliance, we skip coordination
yet continue to schedule follow-ups.
This commit ensures that we bootstrap a new history_uuid when force
allocating a stale primary. A stale primary should never be the source
of an operation-based recovery to another shard which exists before the
forced-allocation.
Closes#26712
Today when checking settings dependencies, we do not check if fallback
settings are present. This means, for example, that if
cluster.remote.*.seeds falls back to search.remote.*.seeds, and
cluster.remote.*.skip_unavailable and search.remote.*.skip_unavailable
depend on cluster.remote.*.seeds, and we have set search.remote.*.seeds
and search.remote.*.skip_unavailable, then validation will fail because
it is expected that cluster.ermote.*.seeds is set here. This commit
addresses this by also checking fallback settings when validating
dependencies. To do this, we adjust the settings exist method to also
check for fallback settings, a case that it was not handling previously.
Adds Request and Reponse classes for accessing lifecycle policies.
Changes existing tests to use these classes where appropriate.
Sets up SPI configuration to allow parsing *Actions from XContent.
Change the logging infrastructure to handle when the node name isn't
available in `elasticsearch.yml`. In that case the node name is not
available until long after logging is configured. The biggest change is
that the node name logging no longer fixed at pattern build time.
Instead it is read from a `SetOnce` on every print. If it is unset it is
printed as `unknown` so we have something that fits in the pattern.
On normal startup we don't log anything until the node name is available
so we never see the `unknown`s.
This endpoint accepts an arbitrary file in the request body and
attempts to determine the structure. If successful it also
proposes mappings that could be used when indexing the file's
contents, and calculates simple statistics for each of the fields
that are useful in the data preparation step prior to configuring
machine learning jobs.
Watcher validates `action.auto_create_index` upon startup. If a user
specifies a pattern that does not contain watcher indices, it raises an
error message to include a list of three indices. However, the indices
are separated by a comma and a space which is not considered in parsing.
With this commit we change the error message string so it does not
contain the additional space thus making it more straightforward to copy
it to the configuration file.
Closes#33369
Relates #33497
Instead of passing DirectoryService which causes yet another dependency
on Store we can just pass in a Directory since we will just call
`DirectoryService#newDirectory()` on it anyway.
This change collapses all metrics aggregations classes into a single package `org.elasticsearch.aggregations.metrics`.
It also restricts the visibility of some classes (aggregators and factories) that should not be used outside of the package.
Relates #22868
Some browsers (eg. Firefox) behave differently when presented with
multiple auth schemes in 'WWW-Authenticate' header. The expected
behavior is that browser select the most secure auth-scheme before
trying others, but Firefox selects the first presented auth scheme and
tries the next ones sequentially. As the browser interpretation is
something that we do not control, we can at least present the auth
schemes in most to least secure order as the server's preference.
This commit modifies the code to collect and sort the auth schemes
presented by most to least secure. The priority of the auth schemes is
fixed, the lower number denoting more secure auth-scheme.
The current order of schemes based on the ES supported auth-scheme is
[Negotiate, Bearer,Basic] and when we add future support for
other schemes we will need to update the code. If need be we will make
this configuration customizable in future.
Unit test to verify the WWW-Authenticate header values are sorted by
server preference as more secure to least secure auth schemes.
Tested with Firefox, Chrome, Internet Explorer 11.
Closes#32699
It is useful to keep track of which version of a policy is currently
being executed by a specific index. For management purposes, it would
also be useful to know at which time the latest version was inserted
so that an audit trail is left for reconciling changes happening in ILM.
This commit allows us to use different TranslogRecoveryRunner when
recovering an engine from its local translog. This change is a
prerequisite for the commit-based rollback PR.
Relates #32867
The main benefit of the upgrade for users is the search optimization for top scored documents when the total hit count is not needed. However this optimization is not activated in this change, there is another issue opened to discuss how it should be integrated smoothly.
Some comments about the change:
* Tests that can produce negative scores have been adapted but we need to forbid them completely: #33309Closes#32899
Daniel Mitterdorfer
Jay Modi
Chris Roberson
Armin Braun
Colin Goodheart-Smithe
Martijn van Groningen
Nhat Nguyen
Tim Vernum
David Roberts
David Turner
Andrei Stefan
Lee Hinman
Nik Everett
Nathan Delhaye
Jason Tedor
Dimitris Athanasiou
Gordon Brown
David Kyle
Albert Zaharovits
Ed Savage
Ioannis Kakavas
Yogesh Gaikwad
Tal Levy
Kazuhiro Sera
Lee Hinman
Jim Ferenczi
jaymode
Marios Trivyzas
Benjamin Trent
Shaunak Kashyap
Ryan Ernst
Hendrik Muhs
Alexander Reelsen
Costin Leau
Christoph Büscher
Tim Brooks
Daniel Mitterdorfer
Simon Willnauer
Tim Vernum
Alan Woodward
markharwood
Or Bin
Ioannis Kakavas
Tanguy Leroux
Vladimir Dolzhenko