Not only was StringJoiner unused, it's also a class only available in java 1.8, which is a problem given that the REST client has minimum java required set to 1.7
The warnings get printed out in a single line e.g. WARNING: request [DELETE http://localhost:9200/index/type/_api] returned 3 warnings:[this is warning number 0],[this is warning number 1],[this is warning number 2]
If you try to close the rest client inside one of its callbacks then
it blocks itself. The thread pool switches the status to one that
requests a shutdown and then waits for the pool to shutdown. When
another thread attempts to honor the shutdown request it waits
for all the threads in the pool to finish what they are working on.
Thus thread a is waiting on thread b while thread b is waiting
on thread a. It isn't quite that simple, but it is close.
Relates to #22027
Changes the default socket and connection timeouts for the rest
client from 10 seconds to the more generous 30 seconds.
Defaults reindex-from-remote to those timeouts and make the
timeouts configurable like so:
```
POST _reindex
{
"source": {
"remote": {
"host": "http://otherhost:9200",
"socket_timeout": "1m",
"connect_timeout": "10s"
},
"index": "source",
"query": {
"match": {
"test": "data"
}
}
},
"dest": {
"index": "dest"
}
}
```
Closes#21707
* Rest client: don't reuse that same HttpAsyncResponseConsumer across multiple retries
Turns out that AbstractAsyncResponseConsumer from apache async http client is stateful and cannot be reused across multiple requests. The failover mechanism was mistakenly reusing that same instance, which can be provided by users, across retries in case nodes are down or return 5xx errors. The downside is that we have to change the signature of two public methods, as HttpAsyncResponseConsumer cannot be provided directly anymore, rather its factory needs to be provided which is going to be used to create one instance of the consumer per request attempt.
Up until now we tested our RestClient against multiple nodes only in a mock environment, where we don't really send http requests. In that scenario we can verify that retries etc. work properly but the interaction with the http client library in a real scenario is different and can catch other problems. With this commit we also add an integration test that sends requests to multiple hosts, and some of them may also get stopped meanwhile. The specific test for pathPrefix was also removed as pathPrefix is now randomly applied by default, hence implicitly tested. Moved also a small test method that checked the validity of the path argument to the unit test RestClientSingleHostTests.
Also increase default buffer limit to 100MB and make it required in default consumer
The default buffer limit used to be 10MB but that proved not to be high enough for scroll requests (see reindex from remote). With this commit we increase the limit to 100MB and make it a bit more visibile in the consumer factory.
It was 10mb and that was causing trouble when folks reindex-from-remoted
with large documents.
We also improve the error reporting so it tells folks to use a smaller
batch size if they hit a buffer size exception. Finally, adds some docs
to reindex-from-remote mentioning the buffer and giving an example of
lowering the size.
Closes#21185
Lucene 6.3 is expected to be released in the next weeks so it'd be good to give
it some integration testing. I had to upgrade randomized-testing too so that
both Lucene and Elasticsearch are on the same version.
This enables the RestClient to send array-based (multi-valued) header values, rather than only sending whatever happened to be the _last_ value of the header.
This enables simple support for proxies (beyond proxy host and proxy port, which is done via the RequestConfig)) to provide a base path in front of all requests performed by the RestClient.
This removes final from the RestClient, Response, and Sniffer classes so that outside code can mock them. Their constructors are already package private, so there's not much that can go wrong.
Consuming the response body to make it part of the exception message means that it may not be readable anymore later, depending on whether the entity is repeatable or not. Turns out that the response body tells a lot about the error itself, and considering that we don't expect bodies to be incredibly big for errors, we can wrap the entity into a BufferedHttpEntity to make it repeatable.
Closes#19622
It can happen that the list of healthy hosts is empty, then we get one from the blacklist. but some other operation might have sneaked in and emptied the blacklist in the meantime, so we have to retry till we manage to get some host, either from the healthy list or from the blacklist.
Throw explicit IllegalStateException in unexpected situations, like where both response and exception are set, or when both are unset. Add unit test for SyncResponseListener.
We throw IOException, which is the exception that is going to be thrown in 99% of the cases. A more generic exception can happen, and if it is a runtime one we just let it bubble up as is, otherwise we wrap it into runtime one so that we don't require to catch Exception everywhere, which seems odd.
Also adjusted javadocs for all performRequest methods
We keep the default async client behaviour like in BasicAsyncResponseConsumer, but we lower the maximum size of the buffer from Integer.MAX_VALUE (2GB) to 10 MB. This way users will realize they are buffering big responses in heap hence they'll know they have to do something about it, either write their own response consumer or increase the buffer size limit by providing their manually creeted instance of HeapBufferedAsyncResponseConsumer (constructor accept a bufferLimit int argument).
Also delayed call to HttpAsyncClient#start so that if something goes wrong while creating the RestClient, the http client threads don't linger. In fact, if the constructor fails it is not possible to call close against the RestClient.
HttpClientConfigCallback#customizeHttpClient now also returns the HttpClientBuilder so it can be completely replaced
RequestConfigCallback#customizeRequestConfig now also returns the HttpClientBuilder so it can be completely replaced
The new method accepts the usual parameters (method, endpoint, params, entity and headers) plus a response listener and an async response consumer. Shortcut methods are also added that don't require params, entity and the async response consumer optional.
There are a few relevant api changes as a consequence of the move to async client that affect sync methods:
- Response doesn't implement Closeable anymore, responses don't need to be closed
- performRequest throws Exception rather than just IOException, as that is the the exception that we get from the FutureCallback#failed method in the async http client
- ssl configuration is a bit simpler, one only needs to call setSSLStrategy from a custom HttpClientConfigCallback, that doesn't end up overridng any other default around connection pooling (it used to happen with the sync client and make ssl configuration more complex)
Relates to #19055
The callback replaces the ability to fully replace the http client instance. By doing that, one used to lose any default that the RestClient had set for the underlying http client. Given that you'd usually override one or two things only, like a couple of timeout values, the ssl factory or the default credentials providers, it is not uder friendly if by doing that users end up replacing the whole http client instance and lose any default set by us.
Users wanting to send a request by providing only its method and endpoint, effectively the only two required arguments, shouldn't need to pass in an empty map and a null entity for the body. While at it we can also add a variant to send requests by specifying only method, endpoint and params, but not body. Headers remain a vararg as last argument, so they can always optionally be provided.
Closes#19312
Nik Everett
javanna
Luca Cavanna
Yuhao Bi
Adrien Grand
Chris Earle
Ryan Ernst