PROBLEM STATEMENT:
For a "strange" reason, all our operations of an API have different "host". We have API like this:
operation 1: GET https://host1:port1/api/resources
operation 2: GET https://host1:port2/api/resources/{id}
operation 3: POST https://host2:port3/api/resources
operation 4: POST https://host2:port4/api/resources/search
If we use Swagger/OpenAPI as it is, it means creating one Swagger/OpenAPI specification per operation, resulting having one swagger-ui page per operation, and then, the need to re-create an index page to list all the operations of an API :-/ which is exactly what we want to avoid.
QUESTIONS:
1/ Does this feature - Override "host" and "basePath" at the "/{path}" level - make sense?
2/ Does someone already try to implement this feature in swagger-ui?
3/ Could/should I propose this kind of change to OpenAPI?
Any other useful remarks/comments are welcome ;-)
Overriding the target server at the path or operation level is now supported in OpenAPI 3.0:
openapi: 3.0.0
servers:
- url: https://my.api.com/v1
paths:
/foo:
# Override the server at path level
servers:
- url: https://another.server:8443/basePath
get: ...
post: ...
/bar:
get:
# Override the server at operation level
servers:
- url: https://some.other.server/v2
post: ...
This is not supported in swagger 2.0 specification. It will be added in the next version though, so no need to add the proposal! See here:
https://github.com/OAI/OpenAPI-Specification/issues/562
Related
I'd like to perform an atomic GET in Redis, and if the value returned is equal to some expected value, I'd like to do a SET, but I want to chain all of this together as one atomic operation. (I'm trying to set a flag that indicates whether any process is writing data to disk, as only one process may be permitted to do so.)
Is it possible to accomplish this with Redis?
I have seen documentation on MULTI operations but I haven't seen conditional operations i MULTI operations. Any suggestions others can offer with this would be greatly appreciated!
You can do both the GET and set operations on the redis server itself using Lua scripts. They're atomic and allow you to add logic too.
I ended up using redlock-py, an implementation of the redlock algorithm that the Redis docs recommend for creating write locks: https://redis.io/topics/distlock. The linked article is fantastic reading for anyone looking to create similar write locks in Redis.
redis-if - lua script for "conditional transactions". More convenient than WATCH + MULTY.
You can pass any combination of conditions & followed commands as json object:
const Redis = require('ioredis')
const redis = new Redis()
redis.defineCommand('transaction', { lua: require('redis-if').script, numberOfKeys: 0 })
await redis.set('custom-state', 'initialized')
await redis.set('custom-counter', 0)
// this call will change state and do another unrelated operation (increment) atomically
let success = await redis.transaction(JSON.stringify({
if: [
// apply changes only if this process has acquired a lock
[ 'initialized', '==', [ 'sget', 'custom-state' ] ]
],
exec: [
[ 'set', 'custom-state', 'finished' ],
[ 'incr', 'custom-counter' ]
]
}))
With this script we removed all custom scripting from our projects.
I came across this post looking for a similar type of function, but I didn't see any options that appealed to me. I opted instead to write a small module in Rust that provides this exact type of operation:
https://github.com/KennethWilke/redis-setif
With this module you would do this via:
SETIF <key> <expected> <new>
HSETIF <key> <field> <expected> <new>
You can do this by SET command, with these 2 arguments, which according to the docs here:
GET - return the old string stored at key, or nil if key did not
exist.
NX - Only set the key if it does not already exist.
Since Redis doesn't execute any command while another command is running - you have the 2 operations in an atomic manner.
I am trying to do load test for zuul version 1.1.2.
However I am keep getting following issue after few a minute for running load test.
Caused by: com.netflix.hystrix.exception.HystrixRuntimeException: book could not acquire a semaphore for execution and no fallback available.
at com.netflix.hystrix.AbstractCommand$21.call(AbstractCommand.java:783) ~[hystrix-core-1.5.3.jar:1.5.3]
My question is how can I increase maxSemaphores via confiugration.
hystrix.command.default.execution.isolation.thread.timeoutInMilliseconds= 20000000
zuul.hystrix.command.default.execution.isolation.strategy= SEMAPHORE
zuul.hystrix.command.default.execution.isolation.semaphore.maxConcurrentRequests= 10
zuul.hystrix.command.default.fallback.isolation.semaphore.maxConcurrentRequests= 10
zuul.semaphore.maxSemaphores=3000
zuul.eureka.book.semaphore.maxSemaphore=30000
I have tried search many option on Intenet but one of those works for me
Please advise
it turns out I am using old version. For later version we could set semaphores at Zuul level. below is an example to set the maxSemaphores 3000 as default for routing to every proxied service
zuul.semaphore.maxSemaphores=3000
The actual property is max-semaphores (this would be with yaml config):
zuul:
semaphore:
#com.netflix.hystrix.exception.HystrixRuntimeException: "microservice" could not acquire a semaphore for execution and no fallback available.
max-semaphores: 2000
In the Tinkerpop or Titan documentation, all operations are based on a sample graph. How to creat a new empty graph to work on?
I am programming in erlang connecting to Tinkergraph, planned to use Titan later in production. There is no erlang driver for both so I am connecting by REST. It is easy to read from graph, but if I want to read from user's input then write into the graph, for example, to create a person named teddy:
screenshot 1
I got those errors. What is the correct way?
Thank you.
Update: For following situation:
23> Newperson=terry.
terry
24> Newperson.
terry
If I want to add this terry, below two will not work. What's the correct way to do it?
screenshot 2
1
TitanGraph titanGraph = TitanFactory.open(config); will open a titan graph without the sample data.
If you have already commited the sample data to your keyspace then you can just change the keyspace defined in your config file.
For example if you are using a cassandra backend you would change storage.cassandra.keyspace=xxxxxx .
You can also clear any keyspace using TitanCleanup.clear(graph);
2
As for the error you are seeing. It looks like you are trying to label your vertex incorrectly. I posted the following and it worked:
{
"gremlin" : "g.addV(label, x).property(y,z)",
"bindings" :
{
"x" : "person",
"y" : "name",
"z" : "Teddy"
}
}
A final note, when you start using Titan 1.0.0 make sure you checkout this section of the tinkerpop docs. Especially make sure to change the channel in the gremlin-server.yaml config to:
channelizer: com.tinkerpop.gremlin.server.channel.HttpChannelizer
Answer to my own question: construct a Body by lists:concat() or ++, then post
It looks like Base CRM has upgraded their API and replaced all of their endpoints/parameters.
Previously I was able to retrieve "Won" deals using this call:
session = BaseCrm::Session.new("<LEGACY_ACCESS_TOKEN>")
session.deals.all(stage: :won, sort_by: :last_activity, sort_order: :desc, page: 1)
This query recently started ignoring my parameters, yet it continued to respond with unfiltered data (that was fun when I realized that was happening).
The new syntax is:
client = BaseCRM::Client.new(access_token: "<YOUR_PERSONAL_ACCESS_TOKEN>")
client.deals.where(organization_id: google.id, hot: true)
yet this does not work:
client.deals.where(stage_name: :won)
client.deals.where(stage_name: "Won")
client.deals.where(stage_id: 8) # specified ID found in Base Docs for "Won"
etc.
I've looked into the most recent updates to the Base CRM Gem as well as the Base CRM API Docs but have not found a solution to searching by specific deal stage.
Has anyone had any luck with the new API and this kind of query?
Is there a way to use the legacy API?
I've left message with Base but I really need to fix this, you know, yesterday.
Thanks for your help!
ADDITIONAL INFO
The legacy API/gem responded with JSON where the v2 API/gem responds with a BaseCRM::Deal object:
$ session.deals.find(123456)
# <BaseCRM::Deal
dropbox_email="dropbox#67890.deals.futuresimple.com",
name="Cool Deal Name",
owner_id=54321,
creator_id=65432,
value=2500,
estimated_close_date=nil,
last_activity_at="2016-04-21T02:29:43Z",
tags=[],
stage_id=84588,
contact_id=098765432,
custom_fields={:"Event Location"=>"New York, NY", :Source=>"Friend"},
last_stage_change_at="2016-04-21T02:08:20Z",
last_stage_change_by_id=559951,
created_at="2016-04-18T22:16:35Z",
id=123456,
updated_at="2016-04-21T02:08:20Z",
organization_id=nil,
hot=false,
currency="USD",
source_id=1466480,
loss_reason_id=nil
>
Checkout stage_id. Is this a bug? According to the Docs stage_id should return an integer between 1 and 10.
I've created a custom Symfony execution filter to manage the integrity of a POST request by extending the sfFilter class. The filter works OK but some times I need to disable it and the fact is the parameter condition specified in the filter definition in filters.yml file(see below) is not working.
# filters.yml
rendering: ~
security: ~
SSL:
class: SSLFilter
condition: %APP_ENABLE_SSLFILTER%
integrity:
class: IntegrityFilter
param:
integrity_header: x-integrity-cnonce
condition: %APP_ENABLE_INTEGRITYCHECK%
cache: ~
execution: ~
As you can see the condition for the IntegrityFilter requires the configuration entry ENABLE_INTEGRITYCHECK to be ON in the app.yml file to execute the filter.
# app.yml
all:
sf_guard_plugin:
algorithm_callable: md5
success_signout_url: /
success_signin_url: /
enable_sslfilter: off
# Enables de IntegrityFilter check
enable_integritycheck: off
Well the thing is that's not working and it's becoming since I use a script to enable/disable some settings along with other tasks and that one isn't disabling. The funny thing is that enabling/disabling works fine for the the SSL Filter shown in both YAML files. Any ideas...?