So, DarkDeviL suggested we increase the TTL on that which is currently 10mins, do you have any recom
So, DarkDeviL suggested we increase the TTL on that which is currently 10mins, do you have any recommendation?
BCR has a 10-month minimum lifetime if I remember right.
ROh no, that is too much 
RIrrelevant in that case haha
Bso if you want a longer TTL and your site is frequently accessed, and your assets are not changing too often, then a longer TTL will be beneficial.
BSorry
B10 hours
Bnot 10 months
BCloudflare Docs
Cache Reserve is a large, persistent data store implemented on top of R2 . By pushing a single button in the dashboard, your website’s cacheable …
EWhich makes sense because for TTLs <10h the regular cache is more than likely going to be fine
Bat leat I got the 10 right 
Bso cloudflared does resolve dns locally but then i get a 502
oh, host header ridiculousness.
oh, host header ridiculousness.
RWell, so our main page is using a cached version with data, however, when a user accesses the page it is routing it through a worker that updates the hydration data so basically an html file is cached and then edited by the worker to increase speed. We have around 40k~ unique users a day in the website. We don't update it everyday obviously, but we can always purge specific pages when we do update as part of our CD.
Would you use cache reserve in this case? Is cache reserve a good solution when using purge? how fast is it affected world wide? If I am not gonna use cache reserve, what time would you put in Edge TTL?
Would you use cache reserve in this case? Is cache reserve a good solution when using purge? how fast is it affected world wide? If I am not gonna use cache reserve, what time would you put in Edge TTL?
RI have the same purge question about Tiered Cache btw
RWow, I gotta say, I already see the cache Hit Ratio increased by an INSANE amount lol after enabling Tiered Cache
RCache hit before enabling it was like 14% (I saw it after enabling so it does add up) for last 24 hours.
If I choose 30mins (which has not passed yet) it is already on 60%+.
If I choose 30mins (which has not passed yet) it is already on 60%+.
BIf the worker is caching a file and then editing it, i feel like it makes sense that it would remain in cache for some time... do those HTML files come up with cf-cache-status HIT?
RYep
BAre they unique filenames after the edit, or does it continuously replace the original html?
RIt replaces the hydration data, so the file overall stays the same, only the data within it is changed.
RI would say it changes a 3% out of the html
RIn a specific part
Ban edited file is considered a new file when saved, so likely it is overwriting the same file with new data
RIt is not saved however
Bi think i misworded that statement
RWe are using worker routes btw
Bgotcha.
Rif I didnt metnion that
BWorker stuff is still a bit new to me so i'm a little out of my league here
Bmy gut tells me if a file is being updated frequently, then it's probably being refreshed in cache so frequently that a long TTL doesn't do it any good.
RThe way it works
client -> cf -> cached page -> api request using worker -> changes -> client
client -> cf -> cached page -> api request using worker -> changes -> client
Bhowever, if it did have a long TTL, then Cache Reserve could store it for the worker to pick up later instead of a fresh origin pull
RI believe thats how it works behind the scenes and is the reason I get cache hits there
RYeah so because of that my other questions were related to purging cache, when purging cache with Tiered Cache enabled, is it instant or does it have it's drawback because Tiered Cache is enabled
BIt's relatively instant, within a minute.
CThe worker is using the cache api to cache, or it's just using fetch to fetch a cached asset?
Just important to know that the cache api in workers is local colo only, no tiered caching or cache reserve
Just important to know that the cache api in workers is local colo only, no tiered caching or cache reserve
ROkay, and with Cache Reserve?
BCloudflare Docs
Cache Reserve is a large, persistent data store implemented on top of R2 . By pushing a single button in the dashboard, your website’s cacheable …
BPurge by URL will wipe cache reserve for that URL.
Purge Everything will force a revalidate
Purge Everything will force a revalidate
RI think I made a mistake in what I said, the worker indeed fetches the page and makes a request to the API to get the updated data at the same time
RI'm trying to understand your second sentence as it's abit confusing at first
Bmeans if the worker is caching via api, it doesn't get pushed to tiered cache or cache reserve
RYou mean if I actively use the worker and call a function to cache?
RWhat it does is literally a fetch request to the same url
CThat's fine then, the fetch API supports tiered caching
CWorkers have a specific cache interface: https://developers.cloudflare.com/workers/runtime-apis/cache/ which is local cache only
Cloudflare Docs
Control reading and writing from the Cloudflare global network cache.
Poh no
Poh no
B?
Pmy website is popular right now and imma need to bite the bullet on workers paid
RThanks Chaika and Brandon. It seems like Cache Reserve could also work for us and might actually be the right fit then. Currently I just extended the cache to 1 hour and will probably work on enabling Cache Reserve next week.
Much appreciated!
Much appreciated!
