[ZBXNEXT-2128] Execution of an item at a specific time Created: 2014 Jan 23 Updated: 2018 Nov 13 Resolved: 2015 Dec 13 |
|
Status: | Closed |
Project: | ZABBIX FEATURE REQUESTS |
Component/s: | Agent (G), API (A), Frontend (F), Server (S) |
Affects Version/s: | None |
Fix Version/s: | 3.0.0alpha3 |
Type: | New Feature Request | Priority: | Major |
Reporter: | Alexei Vladishev | Assignee: | Unassigned |
Resolution: | Fixed | Votes: | 14 |
Labels: | flexibleintervals, interval, item, scheduling | ||
Remaining Estimate: | Not Specified | ||
Time Spent: | Not Specified | ||
Original Estimate: | Not Specified |
Issue Links: |
|
Description |
It would be great if Zabbix supported ability to execute item at a specific time. For example, daily at 09:00, every hour at 0 minutes, etc. Specification https://www.zabbix.org/wiki/Docs/specs/ZBXNEXT-2128 |
Comments |
Comment by richlv [ 2014 Jan 24 ] |
as for syntax, safe way forward would be to use cron syntax |
Comment by Dimitri Bellini [ 2014 Jun 10 ] |
Another suggestion: "Ondemand" execution check. |
Comment by jpka [ 2015 Apr 16 ] |
It looks be great (if not ideal) if new 'Scheduler' menu item can be added to main Zabbix horizontal menu. It can hold and manage items needed for schedule. |
Comment by Andris Zeila [ 2015 Apr 28 ] |
(1) [S] Server part is ready for testing in development branch svn://svn.zabbix.com/branches/dev/ZBXNEXT-2128 <dimir> Great. Please review my changes in r53687, r53886. wiper Thanks, reviewed. |
Comment by Oleksii Zagorskyi [ 2015 Apr 30 ] |
(2) about graphing. Not directly related to current development ! I remembered about this feature development when was writing this comment https://support.zabbix.com/browse/ZBX-7821#comment-142824 Interesting - If I set update interval to 0 and use the new scheduling possibility, for example pool an item every minute or so. Maybe we would need to add new paragraph for example here https://www.zabbix.com/documentation/2.2/manual/config/visualisation/graphs/simple Maybe wort to add those two definitions to https://www.zabbix.com/documentation/2.2/manual/web_interface/definitions because, as I see, they were explained in different issues to several users already. ... discussion is welcome iivs Scheduling intervals do not affect graphs. The documentation here and here could be improved. martins-v see if you can add descriptions there. Ask oleg.egorov if unclear. martins-v I would prefer to deal with this in a separate issue. iivs Then, I guess, we're done here. |
Comment by dimir [ 2015 Jun 01 ] |
Server side tested, front-end part still missing. For now the only way to use it is to edit the database manually according to the spec. The scheduled interval must be written to the delay_flex field, e. g.: update items set delay_flex='wd1-5s/25;wd0-7m/2' where itemid=23661; Both flexible and scheduled intervals may be used separated by semicolon. |
Comment by Bruno Galindro da Costa [ 2015 Jul 07 ] |
This is a nice feature. It will be available in 2.5.0? When 2.5.0 will be available for download? |
Comment by Marc [ 2015 Jul 07 ] |
bruno.galindro, see https://www.zabbix.org/wiki/Docs/roadmap |
Comment by Bruno Galindro da Costa [ 2015 Jul 09 ] |
Tks Marc |
Comment by Andris Zeila [ 2015 Aug 20 ] |
(3) [S] more filter combinations were clarified during frontend development. Server must be tested if they are parsed and processed correctly. Examples: h1-1 m10-20,/5 wiper server accepts out of order intervals (s00h00 for example) while according to the specifications the intervals are listed in 'descending' order. <dimir> It's a bit weird that in case of invalid interval server reports 3 error messages: 14109:20150928:165115.652 Invalid scheduling interval: "wd1m/2h16-18". 14109:20150928:165115.652 wrong delay period format: "wd1m/2h16-18" 14109:20150928:165115.652 wrong delay period format [wd1m/2h16-18] This is with DebugLevel=3. I think one error message would be enough. REOPENED wiper Frontend is also doing interval validation, so those error messages can be generated only if the database was manually altered. The decision was to leave messages, but fix their style. RESOLVED in r55807 <dimir> Right, CLOSED |
Comment by Ivo Kurzemnieks [ 2015 Aug 28 ] |
(4)
Added translation strings:
oleg.egorov CLOSED <richlv> looks like this also removed translatable string "Interval (in sec)" iivs Thanks! Indeed removed. RESOLVED <richlv> thanks for checking this |
Comment by Ivo Kurzemnieks [ 2015 Aug 28 ] |
Frontend ready for testing. |
Comment by Oleg Egorov (Inactive) [ 2015 Sep 03 ] |
(5) Items custom intervals Invalid interval "99999/1-7,00:00-24:00": incorrect syntax near "9/1-7,00:00-24:00". But syntax is ok Other question Scheduling interval Invalid interval "wd1-5h23-22": incorrect syntax near "2". Syntax is correct, but till is less than since. iivs RESOLVED in r55543, r55552 oleg.egorov CLOSED |
Comment by Oleg Egorov (Inactive) [ 2015 Sep 03 ] |
(6) Possible create interval = 0000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000 Missed length/value validation { "itemid": "23663", "delay_flex": "0000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000/1-7,00:00-24:00" } And after this item possible update via frontend Create interval "0" is not allowed, but allow 00...0. I think 0 and 00...0 is same, and possible errors/problems with "0" interval iivs Zero interpretation is fixed in r55502, r55510. About frontend displaying the field with charaters more than maxlength, should be discussed. oleg.egorov Frontend displaying the field with charaters more than maxlength, was discussed and CLOSED. |
Comment by Oleg Egorov (Inactive) [ 2015 Sep 08 ] |
(7) As discussed, please add length validation. iivs Length validation occurrs on database level. oleg.egorov CLOSED |
Comment by Oleg Egorov (Inactive) [ 2015 Sep 08 ] |
(8) Please improve getItemDelay() if (!empty($delay) || !$flexibe_intervals) { !empty($delay) is check for $delay != 0 iivs RESOLVED in r55493 oleg.egorov CLOSED |
Comment by Oleg Egorov (Inactive) [ 2015 Sep 08 ] |
(9) Please review my minor changes in r55487 iivs Some changes were incorrectly changed and some were reverted automatically by merge from trunk. Global coding style issues should be addressed to either oleg.egorov Thanks, please review other minor changes in r55629, r55649, r55692 And as discussed, please rename month_step to month_days_step iivs Renamed and fixed various typos. RESOLVED in r55722 oleg.egorov CLOSED |
Comment by Oleg Egorov (Inactive) [ 2015 Sep 23 ] |
Frontend: TESTED |
Comment by dimir [ 2015 Sep 28 ] |
(10) [F] Delay=0 must be allowed with scheduling intervals as well as it is currently with flexible (spec updated). iivs RESOLVED in r55876 oleg.egorov CLOSED |
Comment by Ivo Kurzemnieks [ 2015 Oct 02 ] |
Implemented in pre-3.0.0alpha3 (trunk) r55922 |
Comment by Andris Zeila [ 2015 Oct 05 ] |
(11) Documentation:
martins-v Updated also:
iivs
martins-v Reviewed. CLOSED. |
Comment by Andris Zeila [ 2015 Oct 08 ] |
(12) [S] If item does not have custom intervals the configuration syncer calls calculate_item_nextcheck() with empty string as custom interval parameter. But when calling calculate_item_nextcheck() from DCget_reachable_nextcheck() NULL is used if item does not have custom interval defined. We should use single approach and better would be to fix configuration syncer so it will also pass NULL as non defined custom interval. wiper RESOLVED in r56014 <dimir> CLOSED |
Comment by dimir [ 2015 Oct 09 ] |
Tested. Please review r56023. wiper Thanks |
Comment by Andris Zeila [ 2015 Oct 09 ] |
Released in:
|
Comment by richlv [ 2015 Oct 22 ] |
(13) removed string "Interval (in sec)" still seems to be present in these test files :
didn't check where exactly these forms were changed, but let's deal with the tests here <richlv> changed in r56434, RESOLVED iivs The string was non-translatable and most of the test files are completely broken anyway. Doesn't really matter if the string is changed in test file or the line is just deleted, this will not repair the test files and they will still be broken. To properly fix this, either old selenium tests should be repaired or new ones should be written. I guess we can leave it as it is for now. CLOSED |
Comment by richlv [ 2015 Oct 29 ] |
subissues still open : |