In the design stage for an app that collects large amounts of data...
Ideally, I want it to be an offline-first app and was looking to Pouchdb/Counchdb - However, the data needs to be kept for years for legal reasons, and my concern is that this is going to consume too much local storage over time.
My thoughts were:
handle sync between pouchdb and couchdb myself, allowing me to purge inactive documents from the local store without impacting the couchdb. This feels messy and probably a lot of work
Build a local store using dexie.js and completely write the sync function. It also looks hard work, but may be less as I'm not trying to mess with a sync function
Search harder :)
Conceptually, I guess I'm looking for a 'DB cache' - holding active json document versions and removing documents that have not been touched for X period. It might be that 'offline' mode is handled separate to the DB cache..
Not sure yet if this is the correct answer..
setup a filter on couchdb to screen out old documents (lets say we have a 'date_modified' field in the doc and we filter out any docs with date_modified older than one month)
have a local routine on the client that deletes documents from the local pouchdb that are older than one month ( actually using the remove() method against the local pouchdb, not updating it with _deleted:true) - from https://pouchdb.com/2015/04/05/filtered-replication.html it appears removed documents don't sync.
docs updated on the Pouchdb will replicate normally
there might be a race condition here for replication, we'll see
This question has been driving me crazy and I can't get my head around it. I come from a MySQL relational background and have been using Meteorjs and Mongo. For the purposes of this question take the example of posts and authors. One Author to Many Posts. I have come up with two ways in which to do this:
Have a single collection of posts - Each post has the author information embedded into the document. This of course leads to denormalization and issues such as if the author name changes how do you keep the data correct.
Have two collections: posts and authors - Each post has an author ID which references the authors collection. I then attempt to do a "join" on a non relational database while trying to maintain reactivity.
It seems to me with MongoDB degrees of denormalization is acceptable and I am tempted to embed as implementing joins really does feel like going against the ideals of Mongo.
Can anyone shed any light on what is the right approach especially in terms of wanting my app data to scale well and be manageable?
Thanks
Denormalisation is useful when you're scaling your application and you notice that some queries are taking too much time to complete. I also noticed that most Mongodb developers tend to forget about data normalisation but that's another topic.
Some developers say things like: "Don't use observe and observeChanges because it's slow". We're building real-time applications so that a normal thing to happen, it's a CPU intensive app design.
In my opinion, you should always aim for a normalised database design and then you have to decide, try and test which fields, that duplicated/denormalised, could improve your app's performance. Example: You remove 1 query per user. The UI need an extra field and it's fast to duplicated it, etc.
With the denormalisation you've an extra price to pay. You've to update the denormalised fields according to the main collection.
Example:
Let's say that you Authors and Articles collections. On each article you have the author name. The author might change his name. With a normalised scenario, it works fine. With a denormalised scenario you have to update the Author document name AND every single article, owned by this author, with the new name.
Keeping a normalised design makes you life easier but denormalisation, eventually, becomes necessary.
From a MeteorJs perspective: With the normalised scenario you're sending data from 2 Collections to the client. With the denormalised scenario, you only send 1 collection. You can also reactively join on the server and send 1 collection to the client, although it increases the RAM usage because of MergeBox on the server.
Denormalisation is something that it's very specify for you application needs. You can use Kadira to find ways of making your application faster. The database design is only 1 factor out of many that you play with when trying to improve performance.
We are investigating using Breeze for field deployment of some tools. The scenario is this -- an auditor will visit sites in the field, where most of the time there will be no -- or very degraded -- internet access. Rather than replicate our SQL database on all the laptops and tablets (if that's even possible), we are hoping to use Breeze to cache the data and then store it locally so it is accessible when there is not a usable connection.
Unfortunately, Breeze seems to choke when caching any significant amount of data. Generally on Chrome it's somewhere between 8 and 13MB worth of entities (as measured by the HTTPResponse headers). This can change a bit depending on how many tabs I have open and such, but I have not been able to move that more than 10%. the error I get is the Chrome tab crashes and tells me to reload. The error is replicable (I download the data in 100K chunks and it fails on the same read every time and works fine if I stop it after the previous read) When I change the page size, it always fails within the same range.
Is this a limitation of Breeze, or Chrome? Or windows? I tried it on Firefox, and it handles even less data before the whole browser crashes. IE fares a little better, but none of them do great.
Looking at performance in task manager, I get the following:
IE goes from 250M memory usage to 1.7G of memory usage during the caching process and caches a total of about 14MB before throwing an out-of-memory error.
Chrome goes from 206B memory usage to about 850M while caching a total of around 9MB
Firefox goes from around 400M to about 750M and manages to cache about 5MB before the whole program crashes.
I can calculate how much will be downloaded with any selection criteria, but I cannot find a way to calculate how much data can be handled by any specific browser instance. This makes using Breeze for offline auditing close to useless.
Has anyone else tackled this problem yet? What are the best approaches to handling something like this. I've thought of several things, but none of them are ideal. Any ideas would be appreciated.
ADDED At Steve Schmitt's request:
Here are some helpful links:
Metadata
Entity Diagram (pdf) (and html and edmx)
The first query, just to populate the tags on the page runs quickly and downloads minimal data:
var query = breeze.EntityQuery
.from("Countries")
.orderBy("Name")
.expand("Regions.Districts.Seasons, Regions.Districts.Sites");
Once the user has select the Sites s/he wishes to cache, the following two queries are kicked off (used to be one query, but I broke it into two hoping it would be less of a burden on resources -- it didn't help). The first query (usually 2-3K entities and about 2MB) runs as expected. Some combination of the predicates listed are used to filter the data.
var qry = breeze.EntityQuery
.from("SeasonClients")
.expand("Client,Group.Site,Season,VSeasonClientCredit")
.orderBy("DistrictId,SeasonId,GroupId,ClientId")
var p = breeze.Predicate("District.Region.CountryId", "==", CountryId);
var p1 = breeze.Predicate("SeasonId", "==", SeasonId);
var p2 = breeze.Predicate("DistrictId", "==", DistrictId);
var p3 = breeze.Predicate("Group.Site.SiteId", "in", SiteIds);
After the first query runs, the second query (below) runs (also using some combination of the predicates listed to filter the data. At about 9MB, it will have about 50K rows to download). When the total download burden between the two queries is between 10MB and 13MB, browsers will crash.
var qry = breeze.EntityQuery
.from("Repayments")
.orderBy('SeasonId,ClientId,RepaymentDate');
var p1 = breeze.Predicate("District.Region.CountryId", "==", CountryId);
var p2 = breeze.Predicate("SeasonId", "==", SeasonId);
var p3 = breeze.Predicate("DistrictId", "==", DistrictId);
var p4 = breeze.Predicate("SiteId", "in", SiteIds);
Thanks for the interest, Steve. You should know that the Entity Relationships are inherited and currently in production supporting the majority of the organization's operations, so as few changes as possible to that would be best. Also, the hope is to grow this from a reporting application to one with which data entry can be done in the field (so, as I understand it, using projections to limit the data wouldn't work).
Thanks for the interest, and let me know if there is anything else you need.
Here are some suggestions based on my experience building on an offline capable web application using breeze. Some or all of these might not make sense for your use cases...
Identify which entity types need to be editable vs which are used to fill drop-downs etc. Load non-editable data using the noTracking query option and cache them in localStorage yourself using JSON.stringify. This avoids the overhead of coercing the data into entities, change tracking, etc. Good candidates for this approach in your model might be entity types like Country, Region, District, Site, etc.
If possible, provide a facility in your application for users to identify which records they want to "take offline". This way you don't need to load and cache everything, which can get quite expensive depending on the number of relationships, entities, properties, etc.
In conjunction with suggestion #2, avoid loading all the editable data at once and avoid using the same EntityManager instance to load each set of data. For example, if the Client entity is something that needs to be editable out in the field without a connection, create a new EntityManager, load a single client (expanding any children that also need to be editable) and cache this data separately from other clients.
Cache the breeze metadata once. When calling exportEntities the includeMetadata argument should be false. More info on this here.
To create new EntityManager instances make use of the createEmptyCopy method.
EDIT:
I want to respond to this comment:
Say I have a client who has bills and payments. That client is in a
group, in a site, in a region, in a country. Are you saying that the
client, payment, and bill information might each have their own EM,
while the location hierarchy might be in a 4th EM with no-tracking?
Then when I refer to them, I wire up the relationships as needed using
LINQs on the different EMs (give me all the bills for customer A, give
me all the payments for customer A)?
It's a bit of a judgement call in terms of deciding how to separate things out. Some of what I'm suggesting might be overkill, it really depends on the amount of data and the way your application is used.
Assuming you don't need to edit groups, sites, regions and countries while offline, the first thing I'd do would be to load the list of groups using the noTracking option and cache them in localStorage for offline use. Then do the same for sites, regions and countries. Keep in mind, entities loaded with the noTracking option aren't cached in the entity manager so you'll need to grab the query result, JSON.stringify it and then call localStorage.setItem. The intent here is to make sure your application always has access to the list of groups, sites, regions, etc so that when you display a form to edit a client entity you'll have the data you need to populate the group, site, region and country select/combobox/dropdown.
Assuming the user has identified the subset of clients they want to work with while offline, I'd then load each of these clients one at a time (including their payment and bill information but not expanding their group, site, region, country) and cache each client+payments+bills set using entityManager.exportEntities. Reasoning here is it doesn't make sense to load several clients plus their payments and bills into the same EntityManager each time you want to edit a particular client. That could be a lot of unnecessary overhead, but again, this is a bit of a judgement call.
#Jeremy's answer was excellent and very helpful, but didn't actually answer the question, which I was starting to think was unanswerable, or at least the wrong question. However #Steve in the comments gave me the most appropriate information for this question.
It is neither Breeze nor the Browser, but rather Knockout. Apparently the knockout wrapper around the breeze entities uses all that memory (at least while loading the entities and in my environment). As described above, Knockout/Breeze would crap out after reading around 5MB of data, causing Chrome to crash with over 1.7GB of memory usage (from a pre-download memory usage around 300MB). Rewriting the app in ANgularJS eliminated the problem. So far I have been able to download over 50MB from the exact same EF6 model into Breeze/Angular, total Chrome memory usage never went above 625MB.
I will be testing larger payloads, but 50 MB more than satisfies my needs for the moment. Thanks everyone for your help.
I'm trying to get my head around the use of IndexDB. I have an SQL database which I access via REST and I'm planning on providing some local caching using IndexDB.
My SQL structure uses a large (and variable) number of tables, each table storing an array of data (time sequence and value) for a specific sensor value. Ideally, I would have assumed I'd create a new object store for each of my tables from MySQL. However, it seems that you can only create an object store when the database is opened which is a bit of a pain.
So, I see a number of options -:
I could use a single object store and add two indexes - one for the time,
and one for the sensor. I'm a little worried that this might have
performance issues, but I'm not sure how data is stored under the
hood.
I could probably detect a new sensor somehow, and open the
database with a new version number. This just feels a little wrong to
me.
I could alternatively use different databases for each sensor,
but I've read somewhere that it's not recommended to use multiple
databases (although it's unclear why since this is possibly the
easiest solution).
I'd welcome any thoughts people have regarding the best structure for this sort of data, that will provide good performance.
If your data sets are independent, example you don't need to combine results from multiple sensors, I suggest you to split them in different tables and/or different databases. Different database option is more convenient for deleting data.
IndexedDB database limit for performance in a single database is for more than 50K data, depending on browser and hardware. I have a couple tests which can measure the speed, just tweak the object size that is inserted and you can test your use case.
If you have less than 10K data per sensor (object store/database) you won't hit big performance issues. One common mistake when inserting batch of data is separate transaction for each insert - this is completely unnecessary, since you can store 10K data with one transaction. If you are working with even larger data set, you can separate the inserting into couple transaction, so you won't block the reading of that database.
Also for every transaction that you do in IndexeDB you need to open a connection, some people use the approach for keeping the one connection alive and reusing it, I prefer the closing and opening a separate connection for each transaction.
Also for faster access, you can store all database info into Local Storage, that way you can track how many databases you have and descriptions for each of them.
Additionally you can take a look at this similar question
I'm interested how does google docs store documents on server side because I need to create similar application.
Does it use pure RTF/ODF files or own database?
How do they make possible versioning and undo/redo feature?
If anybody have knowing according this question please share with me.
To answer you question specifically to how Google Docs works. They use a technology called
Operational Transformation
You may be able to use one of operational transformation engines listed on: https://en.wikipedia.org/wiki/Operational_transform#OT_software
The basic idea is that every operation has a context, e.g. "delete the fourth word in the fifth paragraph" or "add an input box after the button". The clients all send each other operations thru the server. The clients and server each keep their own version of the document and apply operations as they come.
When operations have overlapping contexts, there are a bunch of rules that kick in to resolve conflicts. Like you can't modify something that's been deleted, so the delete must come last in a sequence of concurrent operations on that context.
It's possible that the various clients and server will get out of sync, so you need a secondary algorithm to maintain consistency. One way would be to reload the data from the server whenever a conflict is detected.
--This is an answer I got from a professor when I asked the same thing a couple of years ago.
You should use a database. Perhaps a table storing each document revision. First, find a way to determine whether an update is significant or not. You can store minor changes client side for redo/undo, and then, either periodically or per some condition (e.g., user hits save), create a database entry per revision (you can store things like bytes changed, bytes added, bytes deleted, etc.).
Take a look at MediaWiki, which is open source, and essentially does what you're asking (i.e., take a look at their tables and code).
RTF/ODF would typically be generated, and served, when a user requests exporting the document.
Possibly, you should consider utilizing Google Drive's public API. See link for details.