I'm comparing the service side metrics of my app with the metrics emitted by Cosmos DB and I can see a 100x difference in request counts.
Is my container misconfigured? Am I querying the wrong way? Is Cosmos performing multiple requests internally for each query I'm running against it?
The metric I'm looking at in Cosmos is TotalRequests/Count/5min.
The container has indexes on all attributes + a few composite indexes.
The query I'm running is:
SELECT *
FROM x
WHERE x.partitionKey = 0
and x.index1 = 1
and x.index2 = 2
The container is suffering from a VERY hot partition.
Each request consumes about 5 RUs.
The consistency level is BOUNDED_STALENESS.
I tried changing the consistency level to EVENTUAL which brought the consumed RUs down, but I'm still seeing a huge amount of requests that aren't accounted for.
The Total Requests metric includes every request between the SDK and the service. The SDK makes frequent calls to the service when an SDK instance is first created, then makes regular calls for metadata and other information. If you want to see just requests made by user, apply a filter for OperationType and select the operations you want to monitor.
It's not clear why you were using Bounded Staleness. Reads using Strong and Bounded Staleness consume twice the RU/s because they read from 2 replicas rather than 1 replica for the other weaker consistency models. In addition to differences in cost, there are of course differences in whether you may read stale data or not. They also play a big role in your RTO and RPO in multi-region scenarios.
A hot partition does not have impact on throughput consumption. 5 RU/s for a query is actually very good.
Related
We have implemented Azure Cache Redis for our project.
But the problem is Azure Cache query performance is slower than db query.
For a query while Redis response average is 115ms the db query average is 60ms.
For another query while Redis response average is 200ms db query average is 210ms.
I expected redis queries to return around 50ms for all requests.
Is this normal or are we missing any point.
Maybe speed is not the case all the time?
The performance of AzureRedis cache queries depends on various criteria:
First, you need to check out the source from where you are querying
the Redis cache. If the source and Redis cache resource are in
different regions, there may be a significant network latency.
The pricing tier of the Redis cache also plays a crucial role in the
performance.
Use the redis-benchmark.exe utility to check the throughput and
characteristics of your cache.
You can also consider the scaling link to improve the
performance of the Redis cache.
Possible reasons why query run time varies:
Different data volumes
Different system load
Different query plans
This also can depend on the type of query. In this case SQL database vs Redis.
A website tested this and achieved the following conclusion:
Especially if your request number is large, it's always better to use db and Redis together. Otherwise, both are nearly equal and change in their timings per query.
Related:
Why does the same query takes different amount of time to run?
Redis vs mySQL
We have a solution utilizing a micro-service approach. One of our micro-service is responsible for pushing data to Cosmos. Our Cosmos database is using serverless provision having a 5,000 RU/s limit.
The data we are inserting into Cosmos looks like the below. There are 10 columns and we are pushing a batch containing 5,807 rows of this data.
Id
CompKey
Primary Id
Secondary Id
Type
DateTime
Item
Volume
Price
Fee
1
Veg_Buy
csd2354csd
dfg564dsfg55
Buy
30/08/21
Leek
10
0.75
5.00
2
Veg_Buy
sdf15s1dfd
sdf31sdf654v
Buy
30/08/21
Corn
5
0.48
3.00
We are retrieving data from multiple sources, normalizing it, and sending out the data as one bulk execution to Cosmos. The retrieval process happens every hour. We understand that we are spiking the Cosmos database once per hour with the data that has been retrieved and then stop sending data until the next retrieval cycle. So if this high peak is the problem, what remedies exist for such a scenario?
Can anyone shed some light on what we should/need to do to overcome this issue? Perhaps we are missing a setting when creating the Cosmos database or possibly this has something to do with partitioning?
You can mostly determine these things by looking at the metrics published in the Azure Portal. This doc is a good place to start, Monitor and debug with insights in Azure Cosmos DB.
In particular I would look at the section titled, Determine the throughput consumption by a partition key range
If you are not dealing with a hot partition key you may want to look at options to throttle your writes. This may include modifying your batch size and putting the write operations on a while..loop with a one second timer until RU/s consumed equals 5000 RU/s. You could also possibly look at doing queue-based load leveling and put writes on a queue in front of Cosmos and stream them in.
I am inserting the data to azure cosmos db. In some time it throws an error (Request Timeout : 408). I have increased the Request Timeout to 10 mins.
Also, i have iterate each item from api and calling CreateItemAsync() method instead of bulk executor.
Data To Insert = 430 K Items
Microsoft.Azure.Cosmos SDK used = v3
Container Throughput = 400
Can anyone help me to fix this issue.
Just increase your throughput. But it's going to cost you a lot of money if you leave it increased. 400 RU/s isn't going to cut it unless you batch your operation to the point where it's going to take a long time to insert 400k items.
If this is a one-time deal, increase your RU/s to 2000+, then start slowly inserting items. I would say, depending on the size of your documents, maybe do 50 at a time, then wait 250 milliseconds, then do 50 more until you are done. You will have to play with this though.
Once you are done, move your RU/s back down to 400.
Cosmos DB can be ridiculously expensive, so be careful.
ETA:
This is from some documentation:
Increase throughput: The duration of your data migration depends on the amount of throughput you set up for an individual collection or a set of collections. Be sure to increase the throughput for larger data migrations. After you've completed the migration, decrease the throughput to save costs. For more information about increasing throughput in the Azure portal, see performance levels and pricing tiers in Azure Cosmos DB.
The documentation page for 408 timeouts lists a number of possible causes to investigate.
Aside from addressing the root cause with the SDK client app or increasing throughput, you might also consider leveraging Azure Data Factory to ingest the data as in this example. This assumes your data load is an initialization process and your data can be made available as a blob file.
I have an API which allows other microservices to call on to check whether a particular product exists in the inventory. The API takes in only one parameter which is the ID of the product.
The API is served through API Gateway in Lambda and it simply queries against a Postgres RDS to check for the product ID. If it finds the product, it returns the information about the product in the response. If it doesn't, it just returns an empty response. The SQL is basically this:
SELECT * FROM inventory where expired = false and product_id = request.productId;
However, the problem is that many services are calling this particular API very heavily to check the existence of products. Not only that, the calls often come in bursts. I assume those services loop through a list of product IDs and check for their existence individually, hence the burst.
The number of concurrent calls on the API has resulted in it making many queries to the database. The rate can burst beyond 30 queries per sec and there can be a few hundred thousands of requests to fulfil. The queries are mostly the same, except for the product ID in the where clause. The column has been indexed and it takes an average of only 5-8ms to complete. Still, the connection to the database occasionally time out when the rate gets too high.
I'm using Sequelize as my ORM and the error I get when it time out is SequelizeConnectionAcquireTimeoutError. There is a good chance that the burst rate was too high and it max'ed out the pool too.
Some options I have considered:
Using a cache layer. But I have noticed that, most
of the time, 90% of the product IDs in the requests are not repeated.
This would mean that 90% of the time, it would be a cache miss and it
will still query against the database.
Auto scale up the database. But because the calls are bursty and I don't
know when they may come, the autoscaling won't complete in time to
avoid the time out. Moreover, the query is a very simple select statement and the CPU of the RDS instance hardly crosses 80% during the bursts. So I doubt scaling it would do much too.
What other techniques can I do to avoid the database from being hit hard when the API is getting burst calls which are mostly unique and difficult to cache?
Use cache in the boot time
You can load all necessary columns into an in-memory data storage (redis). Every update in database (cron job) will affect cached data.
Problems: memory overhead of updating cache
Limit db calls
Create a buffer for ids. Store n ids and then make one query for all of them. Or empty the buffer every m seconds!
Problems: client response time extra process for query result
Change your database
Use NoSql database for these data. According to this article and this one, I think choosing NoSql database is a better idea.
Problems: multiple data stores
Start with a covering index to handle your query. You might create an index like this for your table:
CREATE INDEX inv_lkup ON inventory (product_id, expired) INCLUDE (col, col, col);
Mention all the columns in your SELECT in the index, either in the main list of indexed columns or in the INCLUDE clause. Then the DBMS can satisfy your query completely from the index. It's faster.
You could start using AWS lambda throttling to handle this problem. But, for that to work the consumers of your API will need to retry when they get 429 responses. That might be super-inconvenient.
Sorry to say, you may need to stop using lambda. Ordinary web servers have good stuff in them to manage burst workload.
They have an incoming connection (TCP/IP listen) queue. Each new request coming in lands in that queue, where it waits until the server software accept the connection. When the server is busy requests wait in that queue. When there's a high load the requests wait for a bit longer in that queue. In nodejs's case, if you use clustering there's just one of these incoming connection queues, and all the processes in the cluster use it.
The server software you run (to handle your API) has a pool of connections to your DBMS. That pool has a maximum number of connections it it. As your server software handles each request, it awaits a connection from the pool. If no connection is immediately available the request-handling pauses until one is available, then handles it. This too smooths out the requests to the DBMS. (Be aware that each process in a nodejs cluster has its own pool.)
Paradoxically, a smaller DBMS connection pool can improve overall performance, by avoiding too many concurrent SELECTs (or other queries) on the DBMS.
This kind of server configuration can be scaled out: a load balancer will do. So will a server with more cores and more nodejs cluster processes. An elastic load balancer can also add new server VMs when necessary.
I am trying to design create a cloud based system (IaaS) that will gather data from sensors (water pollution related activity) and upon certain events will decide to process the data for a specific sensor.
Data characteristics are:
1. For each sensor data is being sent once every couple of days (up to 6 times a month)
2. each sensor reading contains about 5000 events that are encapsulated in 50-100 messages that are sent to the server (such "session" takes about 20 minutes where messages are sent every 5 seconds)
3. I am building the system to handle rate of 30,000 messages per second.
4. processing of the data shouldn't be real time , I have about 10 minutes once the "session" is finished to do the processing.
5. 90% of the sessions are not interesting and can be thrown away once they are finished. the other 10% have event or event encapsulated in the messages that according to them I need to decide if I need to process the entire session data and send an alert to the sensor that there is a pollution.
I created a tool that generates 5000 messages per second and I am trying to figure out which database would be the most optimal for my scenario.
These are the databases I am thinking to try:
Cassandra - I will save for each session an in memory collection of keys. the keys are for the messages that are stored in cassandra. Once I detect a message that contains bad readings I will need to pull all of the other messages in the "session" and process them (that means 50-100 requests to cassandra). My concern here is about write performance (since I have many read and write operations) + I don't have a good strategy for deleting the 90% not needed sessions.
Couchbase - I will save a document for each "session" according to sensorID and will append each message to the document. Once I detect a message that contains bad readings I will only need to send one request for the document. My concern here is about the read performance.
Redis - use it like cassandra. I assume performance will be the best but I will need to handle the sharding and replication of data myself in order not to reach the memory limit
I would love to hear which option would be the most appropriate
thanks
Reg. Redis – You may consider using a DAAS (Data as a Service). The service will manage for you all the instances, clusters, scaling, data persistence and high availability settings.
One example, is Redis Cloud by Redis Labs
This is an interesting one. If we go to basics of CAP Theorem and try to choose one DB based upon need of consistency, availability, and partition tolerance.
For High consistency and availability- Choose MySQL, PostgreSQL,Greenplum, Vertica, Neo4J.
For High availability and partition tolerance- Use Cassandra,Voldemort,Dynamo,CouchDB, Riak
For High consistency and partition tolerance- Use HBase, Redis, MongoDB,
BerkeleyDB, BigTable
So my Vote is for Cassandra here.