-
Notifications
You must be signed in to change notification settings - Fork 8.2k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Task/hostlist pagination #63722
Task/hostlist pagination #63722
Conversation
Pinging @elastic/endpoint-app-team (Feature:Endpoint) |
Pinging @elastic/endpoint-management (Team:Endpoint Management) |
`); | ||
}); | ||
}); | ||
describe('when an invalid page size is passed', () => { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Are there any other variations you should be testing? Example: what if multiple page_size
params are in the URL? what if page_size
is -1
? or a string (jsdkfjkds
)?
Also - might be worth testing that the parameters get correcly passed down to the middleware/API.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Yes, I agree with @paul-tavares - @parkiino we should test the following invalid values as well:
- -1
- invalid UUID
- valid UUID + special characters
- valid UUID + JS embed
- valid UUID + unicode
- empty string
If it's too much of a lift, I can also test these examples manually and create another ticket to automate these tests in the future. I will also manually test when page_index is more than what we expect to make sure it results in the same bug I found in the policy list URL pagination
@@ -56,7 +56,7 @@ describe('host list middleware', () => { | |||
await sleep(); |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
we should remove the sleeps and instead use the new utility that allows you to wait for a given action to be dispatched - in this case, its probably the serverReturnedHostListData
action.
x-pack/plugins/endpoint/public/applications/endpoint/store/hosts/middleware.ts
Show resolved
Hide resolved
} | ||
if (action.type === 'userChangedUrl' && hasSelectedHost(state) !== false) { | ||
// If user navigated directly to a host details page, load the host list | ||
if (listData(state).length === 0) { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
If a list API call was done, but there were no hosts in the list, then this if()
would alway match. Its unlikely to happen now, but when we introduce the KQL bar, then I can see this being an issue.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
that's true. i'm not sure what to do at that point once the kql bar is introduced. maybe it will be possible to additionally check if filters are active at that point? otherwise, for now i think this makes the most sense for the details page
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
yeah, this would be where perhaps we should track wether the list has been loaded (once) at least. Not a big deal. I guess we can re-evaluate when we introduce filtering
if (listData(state).length === 0) { | ||
const { page_index: pageIndex, page_size: pageSize } = uiQueryParams(state); | ||
try { | ||
const response = await coreStart.http.post('/api/endpoint/metadata', { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I was going to suggest that we add typing to response
by doing:
const response = await coreStart.http.post('/api/endpoint/metadata', { | |
const response = await coreStart.http.post<HostInfo>('/api/endpoint/metadata', { | |
But I just realized that you then get an error because HostInfo
type has been made Imutable (cc/ @oatkiller - maybe making everthing Immutable is the best approach? :) )
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
not sure i'm following. HostInfo
is immutable?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@paul-tavares i think i can type the response, but wouldn't it be HostResultList?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
My mistake - @parkiino you are correct - you should use HostResultList
type here :)
}; | ||
} | ||
} | ||
// otherwise we are not on a host list or details page | ||
return { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
You probably should reset the entire state as in:
return initialState();
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
i know there's something going on to move the location out of all the individual reducers but that hasn't merged yet right? initially i did it that way so that the location would still be updated in the host reducer
history.push( | ||
urlFromQueryParams({ | ||
...queryParams, | ||
page_index: JSON.stringify(index), |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
(optional) You don't need stringify here and below
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
because the type is actually set up as string, typescript complains if i don't have stringify it here. maybe i should just turn that into a number and all will be solved?!
I checked out the pull request and ran it locally. Here is some more feedback:
|
@@ -56,7 +56,7 @@ describe('host list middleware', () => { | |||
await sleep(); | |||
expect(fakeHttpServices.post).toHaveBeenCalledWith('/api/endpoint/metadata', { | |||
body: JSON.stringify({ | |||
paging_properties: [{ page_index: 0 }, { page_size: 10 }], | |||
paging_properties: [{ page_index: '0' }, { page_size: '10' }], |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
the page_index and page_size are strings because they are coming from the url now
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This does not seem valid. The Server Schema for the metadata POST request indicates these should be numbers.
See
page_size: schema.number({ defaultValue: 10, min: 1, max: 10000 }), |
if you are sending strings to the server now (in the browser) and that is succeeding, then perhaps the platform translates it back to a number 🤷♂️
|
||
// if on the host list page for the first time, return new location and load list | ||
if (isCurrentlyOnListPage) { | ||
if (!wasPreviouslyOnListPage) { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
we currently refresh the list if the user closes a host details flyout
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think the logic that decides to set loading
to true
here should be used to decide when to actually make the request.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
as opposed to in the middleware? not really sure i understand what you mean
paging_properties: [{ page_index: pageIndex }, { page_size: pageSize }], | ||
}), | ||
}); | ||
response.request_page_index = Number(pageIndex); |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
What's the purpose of this?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
the api doesn't return the page index that the ui would use so we manually set it in the ui
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Consider typing response
with a type that's shared with the server. If the server should be returning this value, why don't you just change the code there instead of here?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
the server returns what would originally be returned from an elastic search query, which is the page index within elastic i think? or something like that. so it seemed to make more sense to be consistent with that
@@ -88,23 +88,40 @@ export type SubstateMiddlewareFactory = <Substate>( | |||
middleware: ImmutableMiddleware<Substate, AppAction> | |||
) => Middleware<{}, GlobalState, Dispatch<AppAction | Immutable<AppAction>>>; | |||
|
|||
export interface HostListState { | |||
export interface HostState { | |||
/** list of host **/ |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
👍
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
few q's
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@parkiino - can you review these test cases and update the AC for host list pagination? I can document them in TestRail for us. I checked off the ones already in your PR. I'll execute the others manually until we can automate later:
- host list url params uses page_index with an invalid value
- host list url params has a page_index with a value that is more than total pagination
- host list displays correct page when using url params directly
- host list url params accepts only positive numbers for page_index and page_size
- host list url params accepts only a value of 10, 20 or 50 for page_size (what are our defaults?)
- host list url params ignores unknown search params
- host list url params uses only first search param value if multiple values are present
- host list url params uses defaults for any missing search params
Object { | ||
"page_index": "0", | ||
"page_size": "20", | ||
} |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@parkiino - are these values the defaults we are using? It would be good to test the defaults for this test case and have a separate test case for invalid values that revert to defaults.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
the defaults are going to be page_index: 0, and page_size: 10
`); | ||
}); | ||
}); | ||
describe('when an invalid page size is passed', () => { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Yes, I agree with @paul-tavares - @parkiino we should test the following invalid values as well:
- -1
- invalid UUID
- valid UUID + special characters
- valid UUID + JS embed
- valid UUID + unicode
- empty string
If it's too much of a lift, I can also test these examples manually and create another ticket to automate these tests in the future. I will also manually test when page_index is more than what we expect to make sure it results in the same bug I found in the policy list URL pagination
|
@@ -56,7 +56,7 @@ describe('host list middleware', () => { | |||
await sleep(); | |||
expect(fakeHttpServices.post).toHaveBeenCalledWith('/api/endpoint/metadata', { | |||
body: JSON.stringify({ | |||
paging_properties: [{ page_index: 0 }, { page_size: 10 }], | |||
paging_properties: [{ page_index: '0' }, { page_size: '10' }], |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This does not seem valid. The Server Schema for the metadata POST request indicates these should be numbers.
See
page_size: schema.number({ defaultValue: 10, min: 1, max: 10000 }), |
if you are sending strings to the server now (in the browser) and that is succeeding, then perhaps the platform translates it back to a number 🤷♂️
} | ||
if (action.type === 'userChangedUrl' && hasSelectedHost(state) !== false) { | ||
// If user navigated directly to a host details page, load the host list | ||
if (listData(state).length === 0) { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
yeah, this would be where perhaps we should track wether the list has been loaded (once) at least. Not a big deal. I guess we can re-evaluate when we introduce filtering
if (listData(state).length === 0) { | ||
const { page_index: pageIndex, page_size: pageSize } = uiQueryParams(state); | ||
try { | ||
const response = await coreStart.http.post('/api/endpoint/metadata', { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
My mistake - @parkiino you are correct - you should use HostResultList
type here :)
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Looks good @parkiino
💚 Build SucceededHistory
To update your PR or re-run it, just comment with: |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Talked with @oatkiller , he's OK for us to move forward with this, we can address additional issues in another PR
Spoke with Rob, he's OK for us to move forward, we can make sure his questions are answered sufficiently in another PR
…bana into ingest-node-pipeline/open-flyout-create-edit * 'feature/ingest-node-pipelines' of github.com:elastic/kibana: (116 commits) [Ingest Node Pipelines] More lenient treatment of on-failure value (elastic#64411) Report Deletion via UI- functional test (elastic#64031) Bump handlebars dependency from 4.5.3 to 4.7.6 (elastic#64402) [Uptime] Update TLS settings (elastic#64111) [alerting] removes usage of any throughout Alerting Services code (elastic#64161) [CANVAS] Moves notify to a canvas service (elastic#63268) [Canvas] Misc NP Stuff (elastic#63703) update apm index pattern (elastic#64232) Task/hostlist pagination (elastic#63722) [NP] Vega migration (elastic#63849) Move ensureDefaultIndexPattern into data plugin (elastic#63100) [Fleet] Fix agent status count to not include unenrolled agents (elastic#64106) Migrate graph_workspace saved object registration to Kibana platform (elastic#64157) Index pattern management UI -> TypeScript and New Platform Ready (edit_index_pattern) (elastic#64184) [ML] EuiDataGrid ml/transform components. (elastic#63447) [ML] Moving to kibana capabilities (elastic#64057) Move input_control_vis into NP (elastic#63333) remove reference to local application service in graph (elastic#64288) KQL removes leading zero and breaks query (elastic#62748) [FieldFormats] Cleanup: rename IFieldFormatType -> FieldFormatInstanceType (elastic#64193) ...
Summary
**Parent Ticket: https://github.com/elastic/endpoint-app-team/issues/331 **
This PR drives host pagination through the URL.
Additionally addresses some wrapping/truncation problems with the hostname in host details as well as truncates all longer values in the host list so that the list is max one line.
Checklist
Delete any items that are not applicable to this PR.
For maintainers