AraigumiHello, does someone know how can I populate the "info" tab of the problems' dashboard? Or the tags? Basically I want to display for example an owner of the machine so that the operator knows who is concerned and can call the person directly
zabbixHello Araigumi, and welcome to #zabbix. I'm your friendly house bot :) Please be patient when it takes a while for people to reply. See our channel guidelines on https://zabbix.org/wiki/Getting_help#IRC and have a nice stay!
volterTags are populated when there are tags for the trigger.
volterAnd Info is automatically populated with numbers, based on operation failure/success.
volterWhat you want, sounds like "Inventory".
volterThat's 2 clicks away from there (JS menu when clicking on the host name).
zabbixbot[04ZBX-13438] 03Unable to update dashboard/Add widgets 14created by anhdoan (12https://support.zabbix.com/browse/ZBX-13438)
zabbixZBX-13438 Open Updated today -- Unable to update dashboard/Add widgets
Araigumithanks @volter, will make use of inventory then!
volterGlad to help
mino_Hi, is there a big difference between the ubuntu 14 and ubuntu 16 packages? i accidentally installed a 16 package on a 14 host, but it seems to be working without problems
Farsider80Hey all, anyone know what the housekeeper table is actually used for in Zabbix. Mine is currently 8Gb with over 140Million records. But we use mysql partioning so have the internal housekeeper disbaled for trends and history so this seems hudge to me
_ikke_Farsider80: Good question. Afaik, when removing things from zabbix, it's the actual housekeeper that deletes the data
_ikke_instead of everything needing to happen in same request
Farsider80Yeah that makes sence to hand it over but doesn't explain why it gets so big
_ikke_Farsider80: Might be the result of the housekeeper being disabled
Farsider80Out of interest what is the table like for others?
_ikke_almost 7 milion records
Farsider80have 3 seperate zabbix enviroments and agree this one looks to be an outlyer
_ikke_all using partitioning + no housekeeper?
Farsider80Yeah on all enviroments. And rest are under 200Mb
zetherooJust to be certain - a trigger with this expression '.str(CRITICAL,#3)}=1' will only fire off when "CRITICAL" is found in 3 consecutive item checks
glebFarsider80: Creating/deleting stuff frequently?
glebzetheroo: "If more than one value is processed, '1' is returned if there is _at least one_ matching value."
glebzetheroo: Use .count(#3,CRITICAL)=3
Farsider80gleb: Not particularly.
glebFarsider80: This table stores items, triggers and LLD rules to be deleted (together with related history and events).
glebThere is no foreign key in history tables (for insert performance) and when items are deleted their history needs to be deleted "manually". Usually housekeeper takes care of this.
Farsider80Well I just took a punt as have a backup and did a "DELETE FROM housekeeper WHERE field = 'itemid' AND value not IN (SELECT itemd FROM items);"
Farsider80Use similar commands as part of my wider cleanup scripts and now have an empty table
glebLooks like a good decision if you are using partitioning and don't care about deleted item history bloating history tables :)
zetheroogleb: this is what I am filling out on the popup dialog for adding an expression http://tinypic.com/r/34dt0nk/9
zetherooI am using 'Count' instead of 'Time' for the 'Last of (T)' value
volterelastic-- Hello, we implemented something, but we didn't really document it. Have fun!
zabbixKarma for elastic is now -5
_ikke_but elastic is nosql, so it must be good, right?
glebzetheroo: I mean count() trigger function
Farsider80i'll give elastic one thing it robust. I have done my best to upset it over the last year or so and it copped with being abused and unkept everytime so far.
volterIt used to be a lot more fragile in previous years.
_ikke_Farsider80: We have to regularly take care of our ES cluster
volter_ikke_: Which version are you on now?
volter< _ikke_> but elastic is nosql, so it must be good, right?
volterOh dear, this _fucked up_ and unversioned DSL!
volterPiles of incomprehensive JSON bullshit statements
_ikke_volter: let me check
_ikke_volter: 2.3.5 aparently
_ikke_a tad old basically
zetheroogleb: I thought I needed to use .str because I need 'CRITICAL' to be found on the page the URL points to.
glebzetheroo: "There are multiple ways to do things"
volter_ikke_: I ran 2.4 for a while and that was OK. I don't remember if 2.3 was causing me a lot of trouble. But that's not to defend the product by any means.
_ikke_volter: Might be in combination of the product that uses ES
zabbix_ikke_ meant: "volter: Might be in combination with the product that uses ES"
tuxickhmm, there's table hosts_templates but no table templates?
voltertemplates are hosts.
tuxickaah right
volterhosts_templates is probably discovery-related
tuxicknot sure if that was such a good idea but ok
volterOr it's the table that links hosts and templates, more likely.
volterWell, I don't see a big problem with that.
tuxickwas looking for a way to back up templates
tuxicksince i tend to fiddle with them
volterUse the API and/or XML export
zabbixbot[04ZBX-13440] 03Cannot edit map with link which does not have triggers 14created by Alexander Vladishev14/sasha (12https://support.zabbix.com/browse/ZBX-13440)
zabbixZBX-13440 Open Updated today -- Cannot edit map with link which does not have triggers
zabbixbot[04ZBX-13441] 03Zabbix Server stopped 14created by Dmitry Fisher14/f1sher (12https://support.zabbix.com/browse/ZBX-13441)
zabbixZBX-13441 Open Updated today -- Zabbix Server stopped
_ikke_Is it me, or has the font-rendering for graphs become blurry in 3.4?
_ikke_not properly anti-aliased
troffaskycan't tell the difference
_ikke_I don't have any reference anymore, so cannot compare
troffaskymaybe in 5 years will have finished upgrading all our 2.2's
_ikke_We luckily just have one instane
zabbixbot[04ZBX-13442] 03Undefined index in map view 14created by Alexander Vladishev14/sasha (12https://support.zabbix.com/browse/ZBX-13442)
zabbixZBX-13442 Open Updated today -- Undefined index in map view
sfl_hi all, anyone with random snmpv3 issue like in https://www.zabbix.com/forum/showthread.php?p=209114
tuxickwhy is queue under 'administration' tab?
tuxickthere's not much to admin in it
_ikke_tuxick: The menu's are aranged by permission level
tuxickthat kinda explains i suppose :)
zabbixHello pbbunny0801, and welcome to #zabbix. I'm your friendly house bot :) Please be patient when it takes a while for people to reply. See our channel guidelines on https://zabbix.org/wiki/Getting_help#IRC and have a nice stay!
pbbunny0801Has anyone seen somethign like this?
pbbunny0801The system its trying is the main zabbix server
zabbixbot[04ZBX-13443] 03Fatal error in map view 14created by Alexander Vladishev14/sasha (12https://support.zabbix.com/browse/ZBX-13443)
zabbixZBX-13443 Open Updated today -- Fatal error in map view
zetheroois there a way to limit the number of characters shown in the Problems column under Monitoring > Problems?
zabbixbot[04ZBX-13444] 03Unable to upload Template from Zabbix 14created by Sameer Rathod14/sameer.rathod (12https://support.zabbix.com/browse/ZBX-13444)
zabbixZBX-13444 Open Updated today -- Unable to upload Template from Zabbix
zetherooI am using this in the Trigger name to grab all the Item check output: {{ITEM.LASTVALUE}.regsub("([\s\S]*)", "\1")}
zabbixbot[04ZBX-13445] 03Zabbix proxy not able to send data to server  14created by Gurvinder Dadyala14/gurvinder (12https://support.zabbix.com/browse/ZBX-13445)
zabbixZBX-13445 Open Updated today -- Zabbix proxy not able to send data to server
troffaskyi wish multiple event generation could be independent from item poll interval
troffaskysomeone is going to tell me to use escalations...really should look into that some time
voltertroffasky: Escalation are not that difficult. :)
tuxickeasy, just forward alert to some manager
amiritehi all, I need usage of the last() function to return a default value if the key I'm trying to retrieve the last of does not exist
amiritethe fact that some items I use the last function with do not exist will always be the case because the items do not come into existence until a particular event has happened in our application, which may or may not happen
amiriteso if i use the last function in a calculated item across say 100 different items and just one of them hasn't yet been initialized it destroys my entire item
amiriteso is there anyway to use last("key", "default fallback value") or something like that?
volteramirite: I don't understand what the problem would be.
volterThe _item_ most certainly does exist, but the trigger is un-evaluated, due to not having any data.
amiritevolter, the item does not exist -- its created through discovery
amiritediscovery is generated through jmx output
volterThen how can the trigger exist?
amiriteACTION shakes fist
amiritelook, i need a way to default a value of last()
amirite"Cannot evaluate function "last()": item "Production_Aggregate:grpsum[{$CZ_WEBSERVERS_HG}, "jmx[\"com.cloudzilla.web.api:type=MediaStats,mediaId=400,packageName=com_samsung_kidsplay,flow=DP\", \"CLIENT:APP_CRASH_ERROR\"]", "last", 0]" does not exist."
amiritethe item definition:
amiritelast("grpsum[{$CZ_WEBSERVERS_HG}, \"jmx[\\"com.cloudzilla.web.api:type=MediaStats,mediaId=399,packageName=com_sniaf_tfro,flow=DP\\", \\"CLIENT:APP_CRASH_ERROR\\"]\", \"last\", 0]") +
amiritelast("grpsum[{$CZ_WEBSERVERS_HG}, \"jmx[\\"com.cloudzilla.web.api:type=MediaStats,mediaId=400,packageName=com_samsung_kidsplay,flow=DP\\", \\"CLIENT:APP_CRASH_ERROR\\"]\", \"last\", 0]") +
amiritelast("grpsum[{$CZ_WEBSERVERS_HG}, \"jmx[\\"com.cloudzilla.web.api:type=MediaStats,mediaId=401,packageName=com_pulmagames_threedpuzzle,flow=DP\\", \\"CLIENT:APP_CRASH_ERROR\\"]\", \"last\", 0]") +
amiritesee how the one with id 400 is missing? its because the app hasn't ran yet
amiritewhen it runs, it will exist
volterNext time, use a pastesite, please.
volterI still claim that the item exists and that the trigger simply didn't evaluate yet, due to the lack of data.
volterOr is that about the aggregate item?
volterI'm not sure if something can be done about that, in that case.
pbbunny0801anyone familiar with the proxy error Z3005
amiritevolter i'm telling you it does not exist because the item will only exist if the discovery json exists for it, and it doesn't exist for it because the jmx bean itself doesn't exist, and that only exists when the app is played
amiritethat's too bad
pbbunny0801it comes with this error: [Z3005] query failed: [0] unable to open database file
pbbunny0801this is on the proxy, and i'm using sqlite3
amiritei think when i'm not inundated i'm going to secretly start playing with statsd/graphite so i can pitch it to my team as a replacement
volterpbbunny0801: Permissions
pbbunny0801of the sqlite3 file itself? or within the db
pbbunny0801volter: i have 4 proxies, i just checked their permission sets, and they're all the same
pbbunny0801i only have one that's spewing this error
pbbunny0801volter: this is the full error
pbbunny0801[Z3005] query failed: [0] unable to open database file [update httptest set nextcheck=1506471381+delay where httptestid=6]
pbbunny0801strange, it kind of got cleared when I restarted the proxy service...huh...i dont like that :(
volterpbbunny0801: I'll skim the code. Which version?
volterBut I'm fairly sure I am right.
volterTry the following:
voltersu - -s/bin/bash zabbix
volterOr whatever user is running the proxy
volterThen: sqlite3 <file>
volterIf that works, I'm wrong.
volterSELinux could also be in the mix.
joenutsGreetings channel, I have zabbix 3.2.11 configured to monitor a host that's only running during business hours. I've set up a maintenance period ( no data collection ) for when the server is not running, but I get a ...<servername> has been restarted zabbix notification once the maintenance window period ends
joenutstimeline is : server starts at 6AM, maintenance period is 00:00 for period of 7h30m, notification that server is/was restarted at 07:37, and clear at 07:47.
joenutsIt appears the restart item is somehow queued during the maintenance window, and sends out notification following the maintenance ( which defeats the purpose of maintenance window for me ). What am I missing ?
volterjoenuts: What's the exact trigger expression?
joenutsIt's part of the default OS Linux template. Trigger expression = {web16.devvpc.example.com:system.uptime.change(0)}<0
volterI guess that makes it clear _why_ this triggers.
volterBut since that host is special, maybe you just want to disable this trigger or add a temporal constraint of some kind?
joenutsI have run across some forum posts that go in that direction, I believe my understanding of a maintenance period with no data collection is flawed and I'd like some clarity
volterDo you understand what that item/trigger does?
joenutsMy understanding : from 00:00 to 07:30 ( configured maintenance window ) the server could blow up, no big deal, if everything is back to normal before 07:30, when maintenance period ends, no triggers are caught
joenutsif the current value of uptime is less than the previous value of uptime, then trigger ( but this should happen at 06:00, and certainly not be triggered by 07:30 )
volterBut you disabled data collection till 7:30.
joenutswheels are turning. so the value will always be less than the previous value following a restart, once data is gathered.
joenutsSo I guess everyone just adds logic to the expression to compensate for when the monitored system has an expected restart?
joenutseasy enough, thanks for the clarification, cheers
volterI think you can come with something clever, but I'm not in the shape to think of it right now.
volterNo worries, you're welcome!
kallenpHowto change trigger value on SPECIFIED host in zabbix linked templates ? I cannot do it, because it is grayed out :/ Need I clone this template ? Or howto do it ?
troffaskycopy the item and edit the copy
troffaskyor...use a macro as the trigger value and override the macro on the host