I'm a bit new to this world of batch vs stream processing and I'm going back and forth in making a call.
In my case, we have an ELT tool that runs jobs both periodically(with intervals varying from 5 mins to 1 year) as well as on-demand. And this tool generates the data (such customer orders, revenues etc..) in the form of a json file which needs to be stored in a data warehouse. Does it make sense to use any streaming service say kafka/kinesis to transport this kind of data, as those jobs could even run once in 5-10 mins or would that be an overkill in this situation?
Related
I am new to Dataflow and pub-sub tools in GCP.
Need to migrate current on prem process to GCP.
Current Process is as follows:
We have two types of data feeds
Full Feed – its adhoc job – Size of full XML is ~100GB (Single XML – very complex one – Complete data – ETL Job process this xml and load it into ~60 tables)
Separate ETL jobs are there to process full feed. ETL job process
full feed and create load ready files and all tables will be truncate
and re-load.
Delta Feed - Every 30 min need to process delta files(XML files – it will have only changes with in last 30 min)
Source system push XML files in every 30 mins(More than one, file has timestamp), scheduled ETL process will pick all the files which are produced by source system and process all the xml files and create 3 load ready files insert, delete and update for each table
Schedule – ETL Jobs are scheduled to run every 5 min, if current process is running more than 5 min, next run will not trigger until current process completes
Order of the file processing is very important(ETL Job will take care of this). Need to process all the files in sequence.
At the end of ETL process load the load ready files into tables (Mainframe)
I was asked to propose the design to Migrate this to GCP. Need to have two process in GCP as well full and delta. My proposed solution should be handle/suitable for both the feeds.
Initially I thought below design.
Pub/sub -> DataFlow -> mySQL/BigQuery
Then came to know that pub/sub will not give the guarantee to process the files in sequence/order. After doing some research learn that recently google introduced ordering key concept for pub/sub, which will make sure to process the messages in order. In google cloud docs it was mentioned that, this feature is in Beta.
I have two questions:
Whether any one used ordering key concept in pub/sub in production environment. If yes, did you face any challenges while implementing this
Is this design is suitable for the above requirement or is there any better solution in GCP
is there any alternative for DataFlow?
Came to know that pub/sub can handle maximum 10MB size of messages, for us each XML size is more than ~5G.
As was mentioned by #guillaume blaquiere, Beta product launching phase brings some restrictions but they are mostly related to the product support:
At beta, products or features are ready for broader customer testing
and use. Betas are often publicly announced. There are no SLAs or
technical support obligations in a beta release unless otherwise
specified in product terms or the terms of a particular beta program.
The average beta phase lasts about six months.
Commonly, Cloud Pub/Sub message ordering feature works as intended, once you have something for developers attention it is highly appreciated to send a report via Google Issue tracker.
I'm aware of the hugely trafficked sites built in Django or Ruby On Rails. I'm considering one of these frameworks for an application that will be deployed on ONE box and used internally by a company. I'm a noob and I'm wondering how may concurrent users I can support with a response time of under 2 seconds.
Example box spec: Core i5, 8Gb Ram 2.3Ghz. Apache webserver. Postgres DB.
App overview: Simple CRUD operations. Small models of 10-20 fields probably <1K data per record. Relational database schema, around 20 tables.
Example usage scenario: 100 users making a CRUD request every 5 seconds (=20 requests per second). At the same time 2 users uploading a video and one background search process running to identify potentially related data entries.
1) Would a video upload process run outside the GIL once an initial request set it off uploading?
2) For the above system built in Django with the full stack deployed on the box described above, with the usage figures above, should I expect response times <2s? If so what would you estimate my maximum hits per second could be for response time <2s?
3) For the the same scenario with Ruby On Rails, could I expect response times of <2s?
4) Specifically with regards to response times at the above usage levels, would it be significantly better built in Java (play framework) due to JVM support for concurrent processing.
Many thanks
Duncan
We have a large table in BigQuery where the data is streaming in. Each night, we want to run Cloud Dataflow pipeline which processes the last 24 hours of data.
In BigQuery, it's possible to do this using a 'Table Decorator', and specifying the range we want i.e. 24 hours.
Is the same functionality somehow possible in Dataflow when reading from a BQ table?
We've had a look at the 'Windows' documentation for Dataflow, but we can't quite figure if that's what we need. We came up with up with this so far (we want the last 24 hours of data using FixedWindows), but it still tries to read the whole table:
pipeline.apply(BigQueryIO.Read
.named("events-read-from-BQ")
.from("projectid:datasetid.events"))
.apply(Window.<TableRow>into(FixedWindows.of(Duration.standardHours(24))))
.apply(ParDo.of(denormalizationParDo)
.named("events-denormalize")
.withSideInputs(getSideInputs()))
.apply(BigQueryIO.Write
.named("events-write-to-BQ")
.to("projectid:datasetid.events")
.withSchema(getBigQueryTableSchema())
.withWriteDisposition(BigQueryIO.Write.WriteDisposition.WRITE_TRUNCATE) .withCreateDisposition(BigQueryIO.Write.CreateDisposition.CREATE_IF_NEEDED));
Are we on the right track?
Thank you for your question.
At this time, BigQueryIO.Read expects table information in "project:dataset:table" format, so specifying decorators would not work.
Until support for this is in place, you can try the following approaches:
Run a batch stage which extracts the whole bigquery and filters out unnecessary data and process that data. If the table is really big, you may want to fork the data into a separate table if the amount of data read is significantly smaller than the total amount of data.
Use streaming dataflow. For example, you may publish the data onto Pubsub, and create a streaming pipeline with a 24hr window. The streaming pipeline runs continuously, but provides sliding windows vs. daily windows.
Hope this helps
We have an app that uses Cloudant as a remote server. Nevertheless, Cloudant is not completely compatible with TouchDB's continuous replications from previous experience. So our alternative for now is to trigger manually one-shot replications at a fixed frequency. Nevertheless, we would like to know if that approach is going to cost us more money than continuous replications, since continuous replications use longpoll and doesn't need to query the server often. In other words, does one-shot pull replications with Cloudant as the target cost us a GET request?
Thank you,
Paul
I think the issue you refer to is [1].
Cloudant's replication is 100% compatible with CouchDB. In this
instance, TouchDB's logs indicate the iOS network stack passed
on incomplete JSON to TouchDB. It's not clear who was to blame
in this case for the replication failure.
[1] https://github.com/couchbaselabs/TouchDB-iOS/issues/241
For the cost question, a one-shot pull replication will result in a GET to the _changes
feed each time it happens, plus the other requests required to
replicate. This _changes request will be counted as a light
HTTP request against your Cloudant account.
However, whether this works out as more or fewer requests overall
depends on the number of changes coming down from the remote server.
It's also important to remember that the number of _changes calls are very small
relative to the number of other calls involved (e.g., getting the
content of the changes themselves and particularly if there are many
attachments).
While this question is specific to TouchDB, and I mention specific
behaviours of that codebase, this answer deals with the requests involved
in replication between any two systems speaking the CouchDB replication
protocol[2].
[2] http://www.dataprotocols.org/en/latest/couchdb_replication.html
Let's take a contrived example: 1 update per 10 second window to
the source database for the replication, where a TouchDB database
is the target. Let's take a 5 minute poll vs. a continuous replication.
For simplicity of call-counting, let's also take attachments out of the
picture. We'll also assume the device has a constant network connection.
For the continuous case, every 10s TouchDB will receive an update in
the _changes feed. This causes the longpoll connection to close.
TouchDB then runs through the changes, requesting the updates from the
source database; one or more GET requests on the remote server. While
this is happening, TouchDB has to open up another longpoll request
to _changes. So in a five minute period, you'd end up with perhaps
30 calls to _changes, plus all the calls to get documents and record
checkpoints.
Compare this with a one-shot replication every five minutes. You'd
receive notification of the 30 updates in one _changes feed call.
TouchDB implements an optimisation[3] whereby it will call _all_docs
to get updated documents for 1- revs, so you might end up with a single
call to get all 30 documents (not possible in the continuous case as
you've received a single change). Then you've the checkpoint documents
to record. At best fewer than 5 HTTP calls, at most about a third of
the continuous case as you've avoided extra _changes requests.
[3] https://github.com/couchbaselabs/TouchDB-iOS/wiki/Replication-Algorithm#performance
It comes down to the frequency of updates you expect to the source
database. One-shot replication is likely to provide a smoother price
curve as you're in better control of the number of requests you make.
A further question is how often connections will drop because of the
network disconnects which happen regularly with mobile devices.
TouchDB's continuous replications will fire back up each time the
user comes on line (if added via the _replicator database). This is a
further source of unpredictable costs.
However, the benefits from more immediate visibility of changes may
certainly be worth the uncertainty.
I have a program which measures the time a student takes to perform a task.
The program is connected to the Internet and the student with the lowest time gets a prize on the website.
Given that the PC could measure time incorrectly (I've seen it happen) or the student could preload a shared library to intercept the OS calls and give phony times, the timing needs to be performed on the server side.
Currently my logic goes like this:
Client signals server to start.
Server records time and sends ACK back to client.
Client starts task after receiving ACK.
Client finishes task and sends notification to server.
Server records time and calculates elapsed time.
The problem is, this includes the latency from the time the Server sends the ACK to the time the Client receives it + the time the Client finishes to the time the Server receives the message. If the latency is 400ms and the task is 6s long, it makes up a rather large percentage of the total time.
Is there any better way measure the time of the task when the student or the PC may be untrustworthy?
Some competitions require the student or competitor to upload his binary program (or mayb the source code) into your server, which you do control. You could do that. In particular, if it is student code, you probably can require to upload the source code, and restrict to some API, and e.g. run it in a chroot-ed environment.