ExamTopics Logo
Mail Us[email protected]
Menu
Amazon Discussions
exam questions

Exam AWS Certified Big Data - Specialty All Questions

View all questions & answers for the AWS Certified Big Data - Specialty exam
Go to Exam

Exam AWS Certified Big Data - Specialty topic 1 question 48 discussion

Exam question from Amazon's AWS Certified Big Data - Specialty
Question #: 48
Topic #: 1
[All AWS Certified Big Data - Specialty Questions]

An online gaming company uses DynamoDB to store user activity logs and is experiencing throttled writes on the companys DynamoDB table. The company is NOT consuming close to the provisioned capacity. The table contains a large number of items and is partitioned on user and sorted by date. The table is 200GB and is currently provisioned at 10K WCU and 20K RCU.
Which two additional pieces of information are required to determine the cause of the throttling? (Choose two.)

  • A. The structure of any GSIs that have been defined on the table
  • B. CloudWatch data showing consumed and provisioned write capacity when writes are being throttled
  • C. Application-level metrics showing the average item size and peak update rates for each attribute
  • D. The structure of any LSIs that have been defined on the table
  • E. The maximum historical WCU and RCU for the table
Show Suggested Answer Hide Answer
Suggested Answer: AD 🗳️
by Hitu at Sept. 16, 2019, 2:59 a.m.
Disclaimers:
  • - ExamTopics website is not related to, affiliated with, endorsed or authorized by Amazon.
  • - Trademarks, certification & product names are used for reference only and belong to Amazon.

Comments

Chosen Answer:
This is a voting comment (?). It is better to Upvote an existing comment if you don't have anything to add.
Switch to a voting comment New
Submit Cancel
freedomeox
Highly Voted 3 years, 7 months ago
AC in my opinion. The key of this question is to figure out the cause of throttling under capacity. I take the "capacity" as the wcu and rcu you defined when creating a new dynamodb table. the capacity is the TOTAL wcu and rcu in ALL partitions together. the max wcu per partition is 1000, the max size per partition is 10G. However, we have 200G in total and the total wcu requested is 10k. This means, dynamoDB has to give us 20 partitions, with 500 wcu for each partition. therefore, a throttling will occur if for a single partiton the wcu goes over 500. option c gives us information of the most frequently updated attributes and the size, thus can help us calculate whether a wcu for a single partition is over 500. On the other hand, since GSI uses its own capacity, so even the traffic is under "general capacity", it might exceed the GSI capacity, thus we need to know if there is any GSI, and the capacity corresponding to it.
upvoted 9 times
vicks316
3 years, 6 months ago
Spot on, this is exactly my reasoning as well, 100% agree with this.
upvoted 1 times
...
...
hoty
Most Recent 3 years, 6 months ago
A & B are the right answers
upvoted 1 times
...
Royk2020
3 years, 6 months ago
Hot partition issue, A & D, GSI has partition key & LSI will have the table partition key as it leftmost key
upvoted 1 times
...
mmk1
3 years, 6 months ago
A and C in my opinion: A - GSI definitely impacts base table writes if GSI WCU is not provisioned correct. C - for understanding if the columns in GSI get updated frequently and compute required WCUs. best practice is to set the WCU same as base table. Throttling on a GSI affects the base table in different ways, depending on whether the throttling is for read or write activity: When a GSI has insufficient read capacity, the base table isn't affected. When a GSI has insufficient write capacity, write operations won't succeed on the base table or any of its GSIs.
upvoted 2 times
...
kein22190
3 years, 6 months ago
My guess is A and C. After reading all comments.
upvoted 2 times
kein22190
3 years, 6 months ago
Changed my mind to A and E. Previous comments: A is right because: ' If your table uses a global secondary index, then any write to the table also writes to the index. If the many writes are occuring on a single partition key for the index, regardless of how well the table partition key is distributed, the write to the table will be throttled too.' https://www.bluematador.com/docs/troubleshooting/aws-dynamo-throttling E: Consider the following scenario: you dial up the throughput for a table because there’s a sudden spike in traffic or you need the extra throughput to run an expensive scan the extra throughputs cause DynamoDB to increase the no. of partitions you dial down the throughput to previous levels, but now you notice that some requests are throttled even when you have not exceeded the provisioned throughput on the table This happens because there are less read and write throughput units per partition than before due to the increased no. of partitions.
upvoted 1 times
...
...
guruguru
3 years, 6 months ago
B and D. This is likely to be a hot partition issue, need to find out how to confirm it is due to hot partition. A. GSI has its own RCU and WCU, will NOT impact the table's RCU and WCU. C. Average item couldn't tell whether hit hot partition or not. E. Historical doesn't mean you won't have hot partition now.
upvoted 2 times
...
kkyong
3 years, 7 months ago
Excessive throttling is caused by: Hot partitions: throttles are caused by a few partitions in the table that receive more requests than the average partition Not enough capacity: throttles are caused by the table itself not having enough capacity to service requests on many partitions the question ask we to forcus on finding hot partitions . A: GSI has different partition key . D: update rates for the key
upvoted 1 times
kkyong
3 years, 7 months ago
GSI has different capcity
upvoted 1 times
...
vicks316
3 years, 6 months ago
Update rates is C, not D. I agree with AC
upvoted 1 times
...
...
certish
3 years, 7 months ago
I think A is incorrect. GSI will have its own capacity units for both Read & Write. Also, the question is to determine the root cause for WRITE throttling and NOT about under-used provisioned capacity. Which means, E is incorrect as well. In my view, it has to be between B, C & D. I will go with B & C.
upvoted 1 times
vicks316
3 years, 6 months ago
Exactly, because GSI has its own WCU, it could be throttled, don't you agree?
upvoted 1 times
...
...
ub19
3 years, 7 months ago
It is given that the company is not consuming close to the provisioned capacity. Therefore, option B and E is not the choice. This requires to know the assess pattern of table, which can determine hot partition issue. Being said that, option C gives clue about assess pattern and option A gives clue if GSI is there and how it is partitioned. D can give some information but it is more aligned with table partition which we know is user with sorted by date.
upvoted 1 times
vicks316
3 years, 6 months ago
So are you agreeing with A&C ? :)
upvoted 1 times
...
...
Bulti
3 years, 7 months ago
Answer is A and E- For A read this link- > https://www.bluematador.com/docs/troubleshooting/aws-dynamo-throttling and for E -> https://theburningmonk.com/2017/05/beware-of-dilution-of-dynamodb-throughput-due-to-excessive-scaling/
upvoted 4 times
certish
3 years, 7 months ago
@Bulit - I like most of your explanation. However for this one, I just thought of sharing my inputs. I think A is incorrect. GSI will have its own capacity units for both Read & Write. Also, the question is to determine the root cause for WRITE throttling and NOT about under-used provisioned capacity. Which means, E is incorrect as well. In my view, it has to be between B, C & D. I will go with B & C.
upvoted 1 times
vicks316
3 years, 6 months ago
A is correct because GSI uses its own WCU/RCU independent of the table's WCU/RCU which can be throttled since the question says that the table's provisioned throughput is not execeeded (at least it is what I understood from it). C is my 2nd option.
upvoted 1 times
...
...
...
aewis
3 years, 7 months ago
I would go for B and D
upvoted 3 times
piemar
3 years, 7 months ago
why would you go for D?
upvoted 1 times
...
...
san2020
3 years, 7 months ago
my selection AD
upvoted 2 times
...
zhengtoronto
3 years, 7 months ago
A is right because: ' If your table uses a global secondary index, then any write to the table also writes to the index. If the many writes are occuring on a single partition key for the index, regardless of how well the table partition key is distributed, the write to the table will be throttled too.' https://www.bluematador.com/docs/troubleshooting/aws-dynamo-throttling
upvoted 4 times
...
ME2000
3 years, 7 months ago
Answer D and E Why option E ? find the explanation DynamoDB provisioned capacity up and down induces less read and write throughput units per partition than before due to the increased no. of partitions after scale down. https://theburningmonk.com/2017/05/beware-of-dilution-of-dynamodb-throughput-due-to-excessive-scaling/
upvoted 2 times
...
PK1234
3 years, 7 months ago
https://docs.aws.amazon.com/amazondynamodb/latest/developerguide/monitoring-cloudwatch.html You do need cloudwatch logs....
upvoted 1 times
...
am7
3 years, 7 months ago
It should be C and D. C because if we get information if a specific attribute only is getting updated which is causing throttling and D because of hot key partition.
upvoted 1 times
Corram
3 years, 7 months ago
Application level metrics are not useful here since we care about summed reads and writes, hence C is wrong. LSI is only about sort keys not partition keys, therefore D is wrong.
upvoted 1 times
...
...
WWODIN
3 years, 7 months ago
It should be caused by Hot Key (Partition), D should be one of the answers
upvoted 1 times
...
Load full discussion...
Community vote distribution
A (35%)
C (25%)
B (20%)
Other
Most Voted
A voting comment increases the vote count for the chosen answer by one.

Upvoting a comment with a selected answer will also increase the vote count towards that answer by one. So if you see a comment that you already agree with, you can upvote it instead of posting a new comment.

SaveCancel
exam
Someone Bought Contributor Access for:
SY0-701
London, 1 minute ago