This website uses Cookies. Click Accept to agree to our website's cookie use as described in our Privacy Policy. Click Preferences to customize your cookie settings.
Accept
Reject
Preferences
Log in to ask a question
Log in to ask a question

What are the maximum values of spike arrest and quota?

Posted on 07-08-2018 10:36 PM
shrenikkumar-s
Silver 1
Silver 1
Reply posted on --/--/---- --:-- AM

@Abhishek Tiwari

@vinita_dwivedi

While spike arrest & quota can be set to values like:

spike_arrest=250PS(per second)

quota=15000/min(i.e. 250*60 since per minute)

1. I was just wondering what are the maximum values that we can set. How can we mention infinite value for these? Setting 0 is theoretically means infinite(any number) value? If not please let me know how to do that.

2. In such cases when we set it to an infinite value(theoretically), how does it work? Practicality APIGEE(anything for that matter) can't handle infinite transactions, so how will it behave?

We've few issues around spike_arrest & quota while performing LnP tests so need these details to understand the behavior.

These details are not captured in the documentation as far as I know/see. If I'm wrong & these are already documented, then please point me to the location where I can find these details. Otherwise waiting promptly for some responses.

0 2 264
Topic Labels
0 Likes
2 REPLIES 2

Posted on --/--/---- --:-- AM
sidd-harth
Bronze 1
Bronze 1
Reply posted on --/--/---- --:-- AM

Hi @Shrenik Kumar SD, AFAIK Spike Arrest & Quota policies are not used to protect Apigee. These policies are used to protect your backend APIs/targets.

One functionality of Spike Arrest policy is to prevent DoS attacks on your backend targets.

Ex - Let's say that you have a target which can only process 1 call every 10 seconds. In that case, you will be using a Spike Arrest with 6pm(which allows only 1 calls every 10 secs. If there is a second call in the first 10 secs, it will not allow it, thus protecting your target)

Quota policy is used to enforce a restriction on the number of calls a client app/developer can make to your api proxy. One of the usecase is monetization.

IMO not using a Quota/Spike policy will allow anyone with the API Endpoint to make infinite number of calls till your targets stops responding.

0 Likes

Posted on --/--/---- --:-- AM
shrenikkumar-s
Silver 1
Silver 1
In response to sidd-harth
Reply posted on --/--/---- --:-- AM

Yes, I'm aware of the usage of these:). The question about the max values that we can enforce and practically how much is actually possible.

Are you saying APIGEE will allow any no of requests(infinite) through it to the backend? I mean APIGEE also might stop responding at some point of time? I mean if I mention 99999ps it can handle so many in a second?(target handles or not is a different topic)- I don't think so.

Note that as I said I wanted this info in order to analyse performance testing issues not to put infinite spike_arrest/quota.

0 Likes