I have some problems understanding the best concept for my problem.
My architecure is pretty basic. I have a backend with data that can be updated and clients which will load data with some filtes.
I have a backend that has the data in a EHCache.
The data model is pretty basic for example
{
id: string,
startDate: date,
endDate: date,
username: string,
group: string
}
The data can only be modified by another backend service.
When data is modified, added or deleted we have an data update event generated.
The clients are all web clients and have a Spring boot REST Service to fetch the data from the cache.
For the data request the clients sends his own request settings. There are different settings like date and text filter. For example
{
contentFilter: Filter,
startDateFilter: date,
endDateFilter: date
}
The backend use this settings to filter the data from the cache and then sends the response with the filtered data.
When the cache generates an update event every client gets notified by a websocket connection.
And then request the full data with the same request settings as before.
My problem is now that there are many cache updates happening and that the clients can have a lots of data to load if the full dataset is loaded everytime.
For example I have this scenario.
Full dataset in cache: 100 000 rows
Update of rows in cache: 5-10 random rows every 1-5 seconds
Client1 dataset with request filter: 5000 rows
Client2 dataset with request filter: 50 rows
Now everytime the client receives a update notification the client will load the complete dataset (5000 rows) and that every 1-5 seconds. If the update only happens on the same row everytime and the row isn´t loaded by the client because of his filter settings then the client would be loading the data unnecessarily.
I am not sure what would be the best solution to reduce the client updates and increase the performance.
My first thought was to just send the updated line directly with the websocket connection to the clients.
But for that I would have to know if the client "needs" the updated line. If the updates are happening on rows that the clients doesn´t need to load because of the filter settings then I would spam the client with unnecessary updates.
I could add a check on the client side if the id of the updated row is in the loaded dataset but then I would need a separate check if a row is added to the cache instead of an update.
But I am not sure if that is the best practice. And unfortunately I can not find many resources about this topic.
The most efficient things are always the most work, sadly.
I won't claim to be an expert at this kind of thing - on either the implementation(s) available or even the best practices - but I can give some food for thought at least, which may or may not be of help.
My first choice: your first thought.
You have the problem of knowing if the updated item is relevant to the client, due to the filters.
Save the filters for the client whenever they request the full data set!
Row gets updated, check through all the client filters to see if it is relevant to any of them, push out to those it is.
The effort for maintaining that filter cache is minimal (update whenever they change their filters), and you'll also be sending down minimal data to the clients. You also won't be iterating over a large dataset multiple times, just the smaller client set and only for the few rows that have been updated.
Another option:
If you don't go ahead with option 1, option 2 might be to group updates - assuming you have the luxury of not needing immediate, real-time updates.
Instead of telling the clients about every data update, only tell them every x seconds that there might be data waiting for them (might be, you little tease).
I was going to add other options but, to be honest, I don't see why you'd worry about much beyond option 1, maybe with an option 2 addition to reduce traffic if that's an issue.
'Best practice'-wise, sending down multiple FULL datasets to multiple clients multiple times a second is certainly not it.
Sending only the data relevant to each client is a much better solution, and if you can further reduce how much the client even needs to send (i.e. only their filter updates and not have them re-send something you could already have saved) is an added bonus.
Edit:
Ah, stateless server - though it's not really stateless. You're using web sockets, so the server has some kind of state for those connections. It's already stateful so option 1 doesn't really break anything.
If it's to be completely stateless, then you also can't store the updated rows of data, so you can't return those individually. You're back to what you're doing which is a full round-trip and data read + serve.
Option 3, though, if you're semi stateless (don't want to add any metadata to those socket connections) but do hold updated rows: timestamp them and have the clients send the time of their last update along with their filters - you can then return only the updated rows since their last update using their provided filters (timestamp becomes just another filter) (or maybe it is stateless, but the timestamp becomes another filter).
Either way, limiting the updated data back down to the client is the main goal if for nothing else than saving data transfer.
Edit 2:
Sounds like you may need to send two bits of data down (or three if you want to split things even further - makes life easier client-side, I guess):
{
newItems: [{...}, ...],
updatedItems: [{...}, ...],
deletedIds: [1,2...]
}
Yes, when their request for an update comes, you'll have to check through your updated items to see if any are deleted and of relevance to the client's filters, but you can send down a minimal list of ids rather than whole rows that your client can then remove.
Related
I would like to retrieve data from my in-memory H2 database via rest endpoint using Spring and Java8. I have 2 endpoints: one to retrieve data, and second one to add data to database.
How can I achieve something like it is described below in easiest way? I am not sure what solution can be better I thought about JMS Queue or CompletableFuture (if it is possible). It should work for few users, they will call to retrieve data saved under their id number.
Scenario:
User calls rest-endpoint to retrieve data.
If data is present in database then it is retrieved and returned to user.
If data is not present in database then connection is hold for 60 seconds and if during that time something will appear in database (added via endpoint to add new data) then data will be returned.
If data is not present in database and new data won’t appear in 60 seconds then endpoint returns no content.
There were multiple ways of doing that and if requirements is clear then i suggest below two approaches.
Approach 1:
Find and retrieve if data available without waiting.
If data not available set resource id and retrieveTime in header and respond to consumer.
Based resource id you can be ready with data if available.
In this way you can sure about your endpoint service time always consistent and ideally it shouldn't be more than 3 seconds.
Approach 2.
if data not available then out sleep in 60 seconds (not in database connection scope) and then again try with same thread.
Don't need any queue or ansyc process here.
Here your loosing resources and service time would take more.
Apart from other approaches, if your systems using eventing then use eventing approach when there is record persistent then send event to consumer (all database has the feature to send event to source system today).
I’m creating a web application
Frontend - reactjs and backend java.
Frontend and backend communicate with each other via rest.
On UI I show a list of items. And I need to filter them for some params.
Option 1: filter logic is on front end
In this case I just need to make a get call to backend and get all items.
After user choose some filter option filtering is happening on ui.
Pros: for that I don’t need to send data to back end and wait for response. Speed of refreshing the list should be faster.
Cons: If I will need multiple frontend clients. Let’s say a mobile app. Than I need to create filters again on this app too.
Option 2: filter logic is on back end
In this case I get all list items when app is loading. After user changes the filter options I need to send a get request with filters params and wait for response.
After that update a list of items on UI.
Pros: filter logic is written only once.
Cons: Speed probably will be much slower. Because it takes time to send request and get a result back.
Question: Where the filter logic should be? In frontend or in backend? Or maybe what is a best practice?
Filter and limit on the back end. If you had a million records, and a hundred thousand users trying to access those records at the same time, would you really want to send a million records to EVERY user? It'd kill your server and user experience (waiting for a million records to propagate from the back end for every user AND then propagate on the front end would take ages when compared to just getting 20-100 records and then clicking a (pagination) button to retrieve the next 20-100). On top of that, then to filter a million records on the front-end would, again, take a very long time and ultimately not be very practical.
From a real world stand point, most websites have some sort of record limit: Ebay = 50-200 records, Amazon = ~20, Target = ~20... etc. This ensures quick server responses and a smooth user experience for every user.
This depends on the size of your data.
For eg: If you are having a large amount of data, it is better to implement the filter logic on the backend and let the db perform the operations.
In case, you have less amount of data, you can do the filter logic on the front end after getting the data.
Let us understand this by an example.
Suppose you have an entity having 1,00,000 records and you want to show it in a grid.
In this case it is better to get 10 records on every call and show it in a grid.
If you want to perform any filter operation on this, it is better to make a query for the db on the backend and get the results
In case it you have just 1000 records in your entity, it will be beneficial to get all the data and do all the filter operations on the frontend.
Most likely begin with the frontend (unless you're dealing with huge amounts of data):
Implement filtering on the frontend (unless for some reason it's easier to do it on the backend, which I find unlikely).
Iterate until filtering functionality is somewhat stable.
Analyze your traffic, see if it makes sense to put the effort into implementing backend filtering. See what percentage of requests are actually filtered, and what savings you'd be getting from backend filtering.
Implement (or not) backend filtering depending on the results of #3.
As a personal note, the accepted answer is terrible advice:
"If you had a million records, and a hundred thousand users trying to access those records at the same time"; nothing is forcing the hundred thousand users to use filtering, your system should be able to handle that doomsday scenario. Backend filtering should be just an optimization, not a solution.
once you do filtering on the backend you'll probably want to do pagination as well; this is not a trivial feature if you want consistent results.
doing backend filtering is likely to become much more complex than just frontend filtering; you should be aware that you're going to spend a significant amount of time (not only for the initial implementation but also for ongoing maintenance) and ask yourself if it's not premature optimization.
TL/DR: Do wherever is easier for you and don't worry about it until it makes sense to start optimizing.
It depends on the specific requirements of your application, but in my opinion the safer bet would be the back-end.
Considering you need filtering in the first place, I assume you have enough data so that paging through it is required. In this case, you need to have the filtering on the back-end.
Lets say you have a page size of 20. After you apply the filter you would expect to have a page of 20 entities that match that specific filtering criteria in the UI. This can't be achieved if you fetch 20 entities, store them in the front-end and afterwards apply the filter on them.
Also, if you have enough data, fetching all of it in the front-end will be impossible due to memory constraints.
I'm currently developing an application in Java that connects to a MySQL database using JDBC, and displays records in jTable. The application is going to be run by more than one user at a time and I'm trying to implement a way to see if the table has been modified. EG if user one modifies a column such as stock level, and then user two tries to access the same record tries to change it based on level before user one interacts.
At the moment I'm storing the checksum of the table that's being displayed as a variable and when a user tries to modify a record it will do a check whether the stored checksum is the same as the one generated before the edit.
As I'm new to this I'm not sure if this a correct way to do it or not; as I have no experience in this matter.
Calculating the checksum of an entire table seems like a very heavy-handed solution and definitely something that wouldn't scale in the long term. There are multiple ways of handling this but the core theme is to do as little work as possible to ensure that you can scale as the number of users increase. Imagine implementing the checksum based solution on table with million rows continuously updated by hundreds of users!
One of the solutions (which requires minimal re-work) would be to "check" the stock name against which the value is updated. In the background, you'll fire across a query to the table to see if the data for "that particular stock" has been updated after the table was populated. If yes, you can warn the user or mark the updated cell as dirty to indicate that that value has changed. The problem here is that the query won't be fired off till the user tries to save the updated value. Or you could poll the database to avoid that but again hardly an efficient solution.
As a more robust solution, I would recommend using a database which implements native "push notifications" to all the connected clients. Redis is a NoSQL database which comes to mind for this.
Another tried and tested technique would be to forgo direct database connection and use a middleware layer like a messaging queue (e.g. RabbitMQ). Message queues enable design of systems which communicate using message. So for e.g. every update the stock value in the JTable would be sent across as a message to an "update database queue". Once the update is done, a message would be sent across to a "update notification queue" to which all clients would be connected. This will enable all of them to know that the value of a given stock has been updated and act accordingly. The advantage to this solution is that you get to keep your existing stack (Java, MySQL) and can implement notifications without polling the DB and killing it.
Checksum is a way to see if data has changed.
Anyway I would suggest you store a column "last_update_date", this column is supposed to be always updated at every update of the record.
So you juste have to store this date (precision date time) and do the check with that.
You can also add a column version number : a simple counter incremented by 1 at each update.
Note:
You can add a trigger on update for updating last_update_date, it should be 100% reliable, maybe you don't need a trigger if you control all updates.
When using in network communication:
A checksum is a count of the number of bits in a transmission unit
that is included with the unit so that the receiver can check to see
whether the same number of bits arrived. If the counts match, it's
assumed that the complete transmission was received.
So it can be translated to check 2 objects are different, your approach is correct.
I have a 20-column grid with anywhere from 100 to 1,000 rows.
If each cell averages 50 characters, I would estimate that a 1000-row grid would consist of 20x50x1000 characters = 1MB.
The data for this grid has to be returned by the server in one (or more) AJAX requests. The grid is un-editable... it is just a way of representing a lot of information (about human genes, in particular).
I am having a hard time deciding whether I should return this in one AJAX request or several. Do you think this is too much data (1MB) to return in the XML/JSON response of one AJAX request? Is this an anti-pattern? Or does it make sense seeing how all the data is logically part of one grid?
This is more of a design question than anything else. I appreciate any feedback.
Could you not load all the data once using a non-Ajax request and then update only the cells that change via Ajax?
Maybe it would interesting keeping the grid "state" in the server so after each field is edited you send the new contents to the server. That would increase the server usage and bandwidth used, but would make it more responsive when the user sends the "submit" command. This also will allow faster input validation (showing an error message almost after the user has modified a cell, and not half hour later).
As an improvement to be in the safe side, keep in memory (JS memory) a list of "dirty" (modified) fields and reset the value when its related ajax response tells you that the server has ack'ed the ajax call; when the user hits "submit" all fields still dirty are sent again to the server.
That said, as long as you stay away from XML, I do not think it is a load so heavy (of course that will depend of hardware and the concurrent users that you have to service).
I'd suggest you fetch the data in batches. i.e, fetch the first 20-25 rows, or just enough to fill the viewport, then gradually load the next batch as the user scrolls through or when he nears the end of the previous batch. That might make it seamless.
Fetching all the data at once might not be an option considering the number of records you might have.
Furthermore, it's not about too much data, it's about the time it takes to fetch that data. I can guarantee you that (according to how you manipulate the json response), the browser can handle any amount of data.
Scenario Imagine a REST service that returns a list of things (e.g. notifications)
Usage A client will continually poll the REST service. The REST service retrieves records from the database. If records are available, they are converted into JSON and returned to the client. And at the same time, the retrieved records are purged from the DB.
Problem How do you handle the problem if the REST endpoints encounters a problem writing the results back to the client ? By that time, the records have been deleted.
Deleting the records will always be a dangerous proposition. What you could do instead is include a timestamp column on the data. Then have your REST url include a "new since" timestamp. You return all records from that timestamp on.
If the notifications grow to be too large you can always setup an automated task to purge records more than an hour old - or whatever interval works well for you.
It sounds like a strange idea to delete DB records after reading access. Possible problems immediately leap into mind: Network trouble prevent the client reading the data, multiple clients cause each other to see incomplete lists, et.al.
The RESTful apporach might be like this:
Give each notification a specific URI. Allow GET and DELETE on these URIs. The client may trigger the record deletion once it successfully received and processed the notification.
Provide an URI to the collection of current notifications. Serve a list of notification data (ID, URI, timestamp, (optional:) content) upon GET request. Take a look at the Atom protocol for ideas. Optional: Allow POST to add a new notification.
With this approach all reading requests stay simple GETs. You may instrument the usual HTTP caching mechanisms on proxies and clients for performance improvement.
Anyway: Deleting a DB entry is a state change on the server. You must not do this upon a GET request. So POST will be you primary choice. But this does not help you much, since the communication might still not be reliable. And polling qith POSTs smells a lot more like Web-Services than REST.
This could be a possible solution.
The REST service can query the database for a list of notifications.
It marks each item in the list ,by say setting a flag in the database.
It then delivers all this notification records to the client.
If the server has successfully sent the results to the client, all marked records are deleted.
In case of failure, the marked records are unmarked,so that they get delivered during the subsequent polling interval.
I hope you got my point.
We did this with special timestamp paramter.
Requests
Request with timestamp.min, this return all items, and server timestamp;
Request with timestamp from server, return items from hat time stamp, and delete prevoius, return server time stamp;
Please note that we did all this with post. means that virtually we sent command (not query get).