stduolchi, I have some question about the api params. I found there is a limit params in host.get method, but there is no offset or page params. how can I set the start cursor for next limit page?
_ikke_stduolc: sadly, this is not possible
_ikke_stduolc: it's only to limit the amount of results you get
_ikke_stduolc: vote for https://support.zabbix.com/browse/ZBXNEXT-2132
zabbixbot[04ZBX-13118] 03Poor readability in action log if presented data is an array 14created by Miks Kronkalns14/Miks.Kronkalns (12https://support.zabbix.com/browse/ZBX-13118)
zabbixZBX-13118 Open Updated today -- Poor readability in action log if presented data is an array
Richlvawwww, secret comments :)
zabbixZBX-13118 Open Updated today -- Poor readability in action log if presented data is an array
RichlvIngus, there are things which can help to build community, yeah...
KofferHi guys, is it possible that zabbix server itself looses templates on some hosts? I have bunch of prod-databases just lost all of its templates (had to re-apply them manually) - from the triggers view I can see that it has had some history, but the history says that the time the trigger was triggered is "never" and no events were made
_ikke_Koffer: Never heard of that or had that kind of issue
q1xKoffer: you might want to check the audit log
_ikke_q1x: That would only help if you knew that the auditlog would log everything :-/
KofferYeah, makes me worry, currently it's no harm no foul, but we could've slept through a major incident... scares me
volter_ikke_: There's still hope! :D
_ikke_I don't think the server ever touches the configuration of hosts, right?
volter_ikke_: Well, it does with LLD!
volterAnd network discovery
volterBut other than that: Probably not
gleb+ auto-registration
jocke_Hi! I'm trying to get the output of a check to show up in a trigger description. I can place {ITEM.VALUE1} in the description but then it cuts off everything after the first line. However, if I go to 'Latest Data' and check History of the item I can see that the full the output is there. I'm using zabbix 3.0.1. Is there any way to get this to work?
jocke_Also, the same macro works in the email that is sent when the trigger goes off
_ikke_jocke_: Macros are truncated at 20 characters anyway
_ikke_jocke_: no way around it currently
jocke_Ok, but why does it work in the email?
jocke_and history?
_ikke_jocke_: because there it is not truncated
_ikke_but everywhere else it is
jocke_aha! ok, thanks for the help!
_ikke_We asked Zabbix for a way around it, but currently they advise you to apply a patch to the front-end
_ikke_You can find that patch in jira (support.zabbix.com)
jocke_Great! thanks alot. Will check it out
volter_ikke_: Truncation in the frontend _can_ be worked around with a patch available on support.zabbix.com.
_ikke_volter: That's what I said, right?
volterjocke_: By the way, 3.0.13 is current!
volter11:03 < _ikke_> jocke_: Macros are truncated at 20 characters anyway
volter11:03 < _ikke_> jocke_: no way around it currently
_ikke_No built-in way
volterAh, OK!
_ikke_And we are very wary of patching zabbix
_ikke_haven't done it so far
volterIt takes some effort, but I did for quite a while with a certain Git workflow and it was at least maintainable.
_ikke_One of the main reasons we switched from Nagios to Zabbix is because of all the customizations we did, which required continuous development effort to keep up-to-date (which never happened)
volterhttps://support.zabbix.com/browse/ZBXNEXT-2448 for reference
SunflowerHi! A newbie question here... "Job for zabbix-agent.service failed because a configured resource limit was exceeded." - This is due to setenforce being enabled (as I read over the net) and now the question - could turning off the setenforce f*** something else up in that matter?
zabbixbot[04ZBX-13119] 03Non working Empty button in Map constructor 14created by Miks Kronkalns14/Miks.Kronkalns (12https://support.zabbix.com/browse/ZBX-13119)
zabbixZBX-13119 Open Updated today -- Non working Empty button in Map constructor
volterSunflower: Do you understand what SELinux does?
finalbeta_ikke_, the same dev will have to happen on zabbix no?
_ikke_finalbeta: Unless you decide not to do any customizations to the source
_ikke_which is a decission our organization made
_ikke_finalbeta: to be fair, we do maintain a custom dashboard which talks to the API
zabbix_ikke_ meant: "finalbeta: to be honest, we do maintain a custom dashboard which talks to the API"
_ikke_But the API is a fairly stable interface
Sunflower@volter I do, however I am just "scared" since the machinne is prod
_ikke_Do you have a non-prod machine you can try it on first?
Sunflowersadly - not really. Regardless I ran setenforce 0 since I know it will turn it off just for the time not being restarted.
Tichodisabling enforcing mode should not cause anything to stop working, but it will make the server less secure if selinux is configured and enabled there for a reason
SunflowerIt didn't work, however, since sestatus still says "Enabled"
_ikke_can't you just raise the resource limit you are running against?
Tichoyou will likely have to set up selinux policy for zabbix if you want to start using it on a server which is in enforcing mode
Tichoso it's a bit more involved than raising a resource limit
SunflowerOver here the guy states that with turning off the SELinux everything should be alright
alhelp zabbix gurus please. https://pastebin.com/tzk3rMuD there is my zab 3.2 installation config and log files. something is wrong with my snmp.
volterSunflower: Just take a look at the audit log, make sure the policy is up to date and add a custom policy, if necessary and desired.
TichoSunflower: if selinux is enabled on that server for a reason, disabling it is a bad idea, you should do what volter says
TichoSunflower: if on the other hand, selinux is enabled there just because it is enabled by default after installation, or perhaps someone thought it would be cool without thinking more about it, and is really not wanted on the server, then disabling it can be considered
volter(And if so, permanently, not just like this.)
SunflowerRight. You're my captain. Thanks. :) The sad part is that I can't prema disable it, since it requires restart :( I will see what I can do.
volterIf you disable it at runtime and add the configuration file entry to disable it permanently, that's good enough.
alvolter help please
q1x_ikke_: well, if there is something there, at least you know it :) if it's not there, it might have been an API call
q1x...or a massive bug...
_ikke_q1x: in my experience there is a lot that is not covered by the audit log
_ikke_even if I directly query the table
volteral: Please ask a question!
volterOh, there
volteral: A device that doesn't like bulk requests, maybe?
volterAssuming the queried OIDs make sense
aldevices are good. no problems, network is good. no bulk requests. i think its some kind timing?
volterAre you sure about the resulting OIDs and the bulk request mode?
volterIf so, how do you know it?
aldo you see my config at the end ? maby i must change something for tuning?
volterWhich version is that?
volterThe server configuration has little to do with your issue, I assume.
alzabbix-server 3.2 updated from 3.0
volterAlso take a lookt at !t internal
volter!t internal
zabbixbotinternalmonitoring: Internal items can help to figure out how full caches are, how busy internal processes are etc. This can be used to determine how many processes are needed. Zabbix ships with default templates for internal monitoring. See http://blog.zabbix.com/monitoring-how-busy-zabbix-processes-are and https://www.zabbix.com/documentation/3.2/manual/config/items/itemtypes/internal
alit is the other issue. i change houskeeper proceses to max (24) but still 2-3 times a day htere is message about more than 75% busy. what can i do?
_ikke_al: housekepper busy is not an issue
_ikke_afaik, it only is either 100% busy, or 100% busy
_ikke_0% busy I meant in the last one
volterAnd in particular, it's unrelated.
IngusRichlv, how would those direct development related comments made public help?
volterWhy would you hide them?
volterI'm getting bloody emails on every new sprint and story point and status change which are totally pointless.
volterHow much less interesting could actual comments be?
finalbeta_ikke_, ah, I understand. You still have to maintain all the templates/settings. (Some monitoring sollutions simply try to do more themselves)
volterI really appreciated that one ticket which has at least a snapshot of the specification and I think nobody was hurt by making it public.
_ikke_finalbeta: yes, of course, but in Nagios, they actually modified the code itself
Ingusvolter, I see your point about the specs but I can´t comment more on that.
_ikke_and tightly integrated with it
_ikke_Ingus: Would be nice if the community could actually contribute back more (except for just supplying patches
IngusIn what way would you be willing to do that?
IngusI was thinking of something like giving releases or ZBXes for testing in return for something else.
volterIngus: Hehe
glebvolter: :D
volterI'm sure abiscuola will support this move.
zabbixbot[04ZBX-13120] 03/usr/sbin/zabbix_server: alerter #1 started(zbx_log_fatal_info+0x13c) [0x7f6f12f3f6ac] 14created by Anton14/apatsev (12https://support.zabbix.com/browse/ZBX-13120)
zabbixZBX-13120 Open Updated today -- /usr/sbin/zabbix_server: alerter #1 started(zbx_log_fatal_info+0x13c) [0x7f6f12f3f6ac]
IngusI can not promise anything yet but the message I want to leave at this moment is that community effort is important and some new ways of collaboration would be expected sooner or later.
volterComing from your side that is?
volterI'm looking forward to that!
zabbixbot[04ZBX-13121] 03CPU statistics collection is initialized in collector instead of main process 14created by Glebs Ivanovskis14/glebs.ivanovskis (12https://support.zabbix.com/browse/ZBX-13121)
zabbixZBX-13121 Open Updated today -- CPU statistics collection is initialized in collector instead of main process
tuxickwhat's the media.active field about?
tuxicktrying to finally figure out this "no media defined" nonsense
voltertuxick: Did you look it up in the API object documentation?
tuxickoww dammit, forgot that exists
tuxickit's so unusual :)
abiscuolavolter: I always appreciate those kind of moves!
zabbixKarma for abiscuola is now 4
zabbixKarma for bringBackSpecsZabbix is now 82
abiscuolaPersonally (and this is ONLY my opinion), there are soooooo many things a good community could provide
abiscuolaAnd IMHO, we are missing (almost) all of them
tuxickok, i guessed the meaning right, but what i see in database and in web interface disagree
voltertuxick: Like how?
abiscuolavolter: I must come to Austria for the stroh one day or another :D
volterIf you just want to _have_ it, give me your address!
volterThough it's probably easy to mail-order.
tuxickvolter: i only see 4 active, should be ~12
tuxicki was wondering the whole media thing broke due to importing database into newer versions etc
abiscuolavolter: I actually really enjoy Austria itself, so the stroh can be an excuse for coming there :D
RichlvIngus, if that was a sincere question, multiple ways (and not limited to these): a) show attention to detail by developers; b) allow community to chime in early in case of incorrect assumptions or other issues; c) allow community to better understand the intent behind decisions, thus identifying thiings as design decisions or bugs; d) along with proper spec process, improve understanding of the design,
Richlvallowing to utilise the features, provide help here etc
Richlvor, as pidzero put it more eloquently - community vs consumers
tuxickoooh haha
tuxickthat's crazy
tuxick0 - enabled
tuxicknevermind :) but who on earth decided to invert that logic?
RichlvIngus, most of the community members who help here, on the tracker and elsewhere don't do that for rmb or similar things :)
Richlvi suspect that most do that for a combination of altriusm, herd effect and a chance to improve their skills
Richlv(and y'all, feel free to chime in with your own reasons :) )
Richlvzabbix actively excludes, pushes away contributions - and the reasons might be banal, like a desire to improve the attractiveness of some commercial services by removing certain information from the community or hiding issues like mistakes
Richlvi was surprised by the exact opposite of redhat managed projects - let's take ansible and awx
Richlvi went on their irc channel and said "guys, you have a minor documentation bug here, wrong licence here and some other minor bug here"
Richlvit would have been faster for them to just go and fix it, but they walked me through the process of submitting a proper pull request for them
tuxickanyway, on status screen under actions there's 'failures', in which i see empty "details" and status "failed"
Richlvso next time i spotted similar minor issues, i just did another pull request, they merged it - and saved some time in the long run
tuxickand no idea what the hell that's about
Richlvby spending 10 minutes they made me more welcome in the community than zabbix...
Richlvand don't get me wrong, zabbix community is awesome, product is great and true opensource/Free software
Richlvbut we have to be honest - zabbix pushes away the community
tuxickhow so?
Richlvotherwise bright-eyed people come in, try to contribute and leave disappointed
Richlvi can give many, many names and nicknames that came in with great patches and other contributions, knocked on doors & windows for a long time, then gave up
Richlvbut oh my, that's the meetup topic i'm not ready yet to put properly in writing :D
Richlvi'll write an article about it one day. it might be too long for anybody to read it, though.
Richlv<abiscuola> And IMHO, we are missing (almost) all of them
_ikke_How is it possible that zabbix says that some items of the same discovery are not discovered anymore, while others remain? (ie, the same single discovery line)
Richlv_ikke_, line ?
Richlvand how - well, some might be still present in json, some might be missing
Richlvif i got you right
Richlvalso :
Richlv!t rmb
zabbixbotRMB: Rīgas melnais balzams - Riga Black Balsam. Alcoholic drink in ceramic bottles, Riga specialty. Taken as a souvenir from trips to Riga
_ikke_It's SNMP, different prototype items with the same index
abiscuolaRichlv: what I noticed, is that there is not an organized process for "handling" a distributed community
Richlv_ikke_, oh. do they use only one lld macro or several ?
abiscuolaIt happened that I handled patches coming from users
Richlvabiscuola, you mean at zabbix or elsewhere ?
abiscuolaAt zabbix
Richlvi'd say that might be intentional to keep full copyright
Richlvwhich is perfectly fine, btw
_ikke_Richlv: ah, that's it, forgot to update the SNMP macro in the key
abiscuolaBah, a lot of projects
Richlv_that_ reason can be handled with care and without isues
Richlv_ikke_, ah, grat it was just that
abiscuolaAsk you to give the copyright to the project/company, so....
abiscuolaI don't think it's a problem for someone that _want_ to contribute
zabbixKarma for Richlv is now 175
abiscuolaGoing back, I said that I handled patches from users
abiscuolaand even if not perfect
abiscuolaI took the time to not only make them better
abiscuolabut also suggest the user what was wrong so they can do "better" next time
abiscuolathat should be a starting point for having a good community
abiscuolause the contributors patches not only as free work, but as teaching moment
abiscuolaAfter that, how many zabbix side projects from the community are out there?
Richlvabiscuola, that absolutely is one of the important pathways. abiscuola++ for that.
zabbixKarma for abiscuola is now 5
zabbixKarma for abiscuola is now 6
abiscuolaand I can safely say that I noticed one single of those
Richlvabiscuola, ...1 ? i'm kidding, i'm kidding - 3.
abiscuolaUnless you go hunt for those
Richlvabiscuola, also, by teaching you get more free work of higher quality. but that's community building 101 :)
abiscuolaRichlv: yep
abiscuolaMaybe this is an argument to discuss with a beer at hand
Richlvabiscuola, i'm considering a revival for an "all things open" meetup
Richlvone thing like zabbix is unlikely to attract attention for long enough
zabbixKarma for Richlv is now 176
Richlvbut if we make it zabbix this time, osm next time, linux-something then, it might just live
Richlvsomething... every two months or so
Richlvdon't think i could make it more frequently myself :)
abiscuolaThat's a good idea
Richlvwe had one years ago, but it fizzled
Richlv(and it was osm, mostly)
_ikke_Any resolution when I changed the item prototype key and it now conflicts with the old items that will be deleted?
Richlv_ikke_, hmm, how does it coflict ? got an example ?
_ikke_Should I revert back first, make sure the items are discovered, and then change it again?
_ikke_Items were created with ifInOctets[{#SNMPVALUE}] item prototypes
_ikke_I caused them to be not discovered anymore (so interfaces says they will be deleted after x amount of hours)
_ikke_I changed the key to ifInOctets[{#IFNAME}]
_ikke_now the discovery rule is not supported anymore due to duplicate items
Richlvwhich version of zabbix ?
Richlvi vaguely remember something like this being fixed/improved
Richlvmaybe check the changelog
Richlvas 3.2.3 is pretty old by now
_ikke_yeah, we plan to upgrade first to 3.2.10 next week, and 3.4 later
Richlvyeah, 3.2 running out of support any moment, too
_ikke_We have been planning to upgrade to 3.4 already, but it takes a bit longer to execute
AlexeyXHi there!
AlexeyXHow to send message about problem in zabbix if the problem duration more than 60 sec....
AlexeyXTherefore, exclude the messages such as - Problem, Ok, Problem, Ok and etc...
AlexeyXwhere time between messages 1 sec... (
RichlvAlexeyX, the most common and easiest option - use trigger functions for that
Richlv!t triggerfu
zabbixbottriggerfunctions: Keywords that tell how to evaluate the data in trigger expressions. Popular functions include last(), min(), max(), avg(), nodata() and count(). See https://www.zabbix.com/documentation/3.2/manual/appendix/triggers/functions
Richlvso instead of last() use things like min(), max(), avg() etc
Richlvbut - 1 second ? what's your item interval ?
AlexeyXit was for example... I will read about it
AlexeyXthank you!
_ikke_Richlv: Cannot find anything related in the release notes (grepped for discovery)
_ikke_Can you have different discovery macros with the same OID?
Richlv_ikke_, it could have been refereed to as lld, LLD or some other way even. also, release notes or changelog ?
Richlv_ikke_, didn't get the last question
_ikke_release notes
Richlvbut probably yes
Richlv_ikke_, i'd suggest changelog
Richlvrelease notes, as far as i know, are webpages
Richlvharder to grep ;)
_ikke_discovery[{#FOO}, 1.2.3, {#BAR}, 1.2.3]
_ikke_Richlv: with the changelog you mean the one in SVN?
Richlv_ikke_, yep
_ikke_Richlv: This seems close: https://support.zabbix.com/browse/ZBX-12259, but I do get an error in the interface
zabbixbot[ZBX-12259] LLD: Multiple OIDs SNMP discovery : Issues if not all OIDs are available on the device (URL: https://support.zabbix.com/browse/ZBX-12259)
Richlvnot sure, haven't tried that myself
troffaskyis there a ZBXNEXT for "Use this address for all interfaces" when setting a host address?
Richlvtroffasky, similar to firefox's proxy setup ?
troffaskyi am talking about Configuration > Hosts > host
troffaskylist of interfaces in there
troffasky90% of hosts I add, I am having to add same information in Agent interfaces and SNMP interfaces
Richlvyeah, so functionally similar to the proxy config in firefox
Richlvthere might, but it's not such a common task, i guess. people with many hosts automate that anyway
troffaskyahhh i know what you mean about firefox now
troffaskyyeah, that
troffaskythe fact that it's not common suggests to me that it needs to default to doing so
troffaskywhilst leaving an option there for whatever use case requires different addresses for different protocols
Richlvthere might be a feature request, do search on jira
Richlvi recall this being discussed, but don't recall the venue
troffaskyyeah, can't find it
troffaskywhilst we're at it, why are IP and name separate fields?
zabbixbot[04ZBXNEXT-4259] 03Use same address for all interfaces 14created by alex dekker14/troffasky (12https://support.zabbix.com/browse/ZBXNEXT-4259)
zabbixZBXNEXT-4259 Open Updated today -- Use same address for all interfaces
_ikke_troffasky: we record both the IP and DNS in there
troffaskyyou are using it as a kind of database, or ...?
_ikke_We monitor mainly by IP
_ikke_but still enter the dns name
_ikke_especially some interfaces have a different dns name as the host
troffaskyas always, it should be optional but zabbix should default to the sanest/commonest use case
volterI used to have a check for whether things still matched.
voltertroffasky: How does it bother you?
troffaskyi want computers to do the work for me
troffaskyzabbix should just work out whether you've entered a name or an IP
troffaskynot challenge me to waste my limited mental bandwidth on it
Richlvtroffasky, it is optional
_ikke_Richlv: discovery issue is already solved again
Richlvthe dns, that is
Richlv_ikke_, yay
troffaskyby optional, i mean the option to specify each separately and specify which to use
Richlvip/dns choice ?
birdmanHi there. I am experiencing an issue interpreting traps from an application. It is: I am receiving the traps, and also watching on the lastest data the notification but it is not informing through the triggers in frontend neither can see what exactly it is happening... Can someone suuggest for me what to do?
Richlvbirdman, please be more specific. for example, do you get data ? what have you configured ?
birdmanYes I get data. I configured the hosts, application templates, and value mappings
birdmanThe problem is while I receive a trap notifying a faiulre iin the monitored host it is supposed to show a problem in frontend but it is not doing
birdmanI just know I receive the data becasuse I see it in the latest data of the host
Richlvwhat triggers do you have ?
birdmanIt says it is receiving a notificacion as trap, but it is not reflected in problems of the host
zabbixbot[04ZBXNEXT-4260] 03Merge name/IP fields on interfaces 14created by alex dekker14/troffasky (12https://support.zabbix.com/browse/ZBXNEXT-4260)
zabbixZBXNEXT-4260 Open Updated today -- Merge name/IP fields on interfaces
birdmanThe trigger I get after parsing the MIB with a process someone shared in github
Richlvtroffasky, 'frontend' label is redundant ;)
zabbixbot[04ZBX-13122] 03Trigger dependancy Issue 14created by Wolfgang Alper14/wolfgang.alper (12https://support.zabbix.com/browse/ZBX-13122)
zabbixZBX-13122 Open Updated today -- Trigger dependancy Issue
Richlvha. wolfgang movet it quickly ;)
Richlvalso typo. "dependency".
SunflowerHey! :) A few hours ago I asked if someone knows how to go around - "Job for zabbix-agent.service failed because a configured resource limit was exceeded". As it was suggested to turn off SELinux over the net here I had the same suggestion. Now however I have SELinux under "Permissive" and the issue persists and I can't start the agent... Anyone got a clue?
finalbetaDid you reboot after changing to permissive?
finalbetaDoes it give the same problem is you start the service manually (directory from cli, not through the service)
RichlvSunflower, sounds more like a systemd issue
Richlvhow did you install the agent ?
Richlva quick search suggests this is often related to files being unreachable, maybe even the pidfile
zabbixbot[04ZBXNEXT-4261] 03Expand the support for macros (HOST.DESCRIPTION,TRIGGER.HOSTGROUP.NAME,INVENTORY.CONTACT) 14created by Aleksejs Petrovs14/aleksejs.petrovs (12https://support.zabbix.com/browse/ZBXNEXT-4261)
Sunflowersorry for not answering, I was reading the logs. Precisely, journalctl is saying - [hostname] PID file /run/zabbix/zabbix_agentd.pid not readable (yet?) after start. [hostname] Failed to start Zabbix Agent.
SunflowerAnd Rich - I installed the agent trough adding the repo and then yum install (its centos 7 btw)
troffaskyhow long after editing "Keep lost resources period" would the "will be deleted in ..." be re-evaluated?
zabbixbot[04ZBX-13123] 03Add php-mysqli optional dependency to Debian packages 14created by Toni Viemerö14/todeveni (12https://support.zabbix.com/browse/ZBX-13123)
zabbixZBX-13123 Open Updated today -- Add php-mysqli optional dependency to Debian packages
zabbixbot[04ZBX-13124] 03Unable to execute remote command on agent after host autoregistration 14created by Kaspars Mednis14/kaspars.mednis (12https://support.zabbix.com/browse/ZBX-13124)
RichlvSunflower, uh-oh, a common problem. volter, do you recall what was the solution when somebody had it recently here ?
zabbixZBX-13124 Open Updated today -- Unable to execute remote command on agent after host autoregistration
Richlvtroffasky, the mesages are updated the next time the discovery is run
Richlvwell, or more like not messages, but the fact that it's not discovered
Richlvthere's definitely at least one older issue about https://support.zabbix.com/browse/ZBX-13124, but i'll let somebody else search for it
troffaskyok, will give it an hour then
zabbixZBX-13124 Open Updated today -- Unable to execute remote command on agent after host autoregistration
Richlvwas discussed quite a bit
volterRichlv: Having a different pidfile set in the configuration?
volterSomebody also claimed once that /run was not a symlink to /var/run in his case or something along those lines.
volterOther than that the directory in /run might not exist or permissions may not allowed to write to it.
volterWhich would indicate a packaging error or manual intervention.
volterSunflower: ^
Richlvyeah, it seems to all blow up every other day for somebody
Richlvsystemd's baby issues will likely be for a decade )
volterThe way the unit file is crafted doesn't really help it.
volterThis is not a systemd issue, but you know that.
volterYou must simply stop to apply your SysV logic to everything.
Richlvi like my sysv logic (and i save my exclamation marks)
Richlvstumbled upon https://english.stackexchange.com/questions/11607/when-and-how-should-i-use-multiple-exclamation-marks
Richlv'Multiple exclamation marks,' he went on, shaking his head, 'are a sure sign of a diseased mind.'
troffaskyi agree!!!
amiritehow do I restrict what certs I accept as a zabbix server? TLSAccept doesn't seem to be a parameter for zabbix_server.conf
amiritehow do I enforce encryption for incoming connections to the server?
amiritesomeone please tell me the secret
paulwasianyone have insight into why the trends and trends_uint tables don't have indexes especially including the clock column?
roycrofthello, folks
zabbixHello roycroft, 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!
roycrofti'm doing a new zabbix install - we've been working with really old versions on really old operating systems, and are close to deploying it in production, so i want to get things up-to-date
zabbixbot[04ZBX-13125] 03not working flags g modifier in regexp 14created by Sergey Korchak14/malmstream (12https://support.zabbix.com/browse/ZBX-13125)
zabbixZBX-13125 Open Updated today -- not working flags g modifier in regexp
roycroftso i built a vm running debian stetch, and did apt-get install zabbix-mysql-server zabbix-frontend-php
roycroftthat all went fine, except zabbix is insisting that the only database engine it will use is postgresql
roycroftpersonally, i would prefer that, but my boss insists that we run it on mariadb
roycroftdoes anyone know what's going on?
roycroftis the debian stretch zabbix-mysql-server package broken?
roycroftor is the zabbix-frontend-php package broken?
_NiCroycroft, which repo did you install from?
volterroycroft: You are missing the mysql PHP module.
volterAssuming that's the frontend claiming this.
volterpaulwasi: The combined primary key will create an index.
zabbixbot[04ZBX-13126] 03not working g modifier in regexp 14created by Sergey Korchak14/malmstream (12https://support.zabbix.com/browse/ZBX-13126)
zabbixZBX-13126 Open Updated today -- not working g modifier in regexp
paulwasi<volter> oh I see thank you. I'm used to the mssql gui tool showing you multicolumn primary key indexes in the indexes view
paulwasithe pgAdmin 4 tool omits them in case anyone was wondering where I was missing it
morgajelweird question: I'm seeing a lot of the following in my proxy logs: cannot send list of active checks to "someipaddress": host [somehostname] not found
morgajelI searched for both the hostname and the IP in zabbix proper and don't see any entries (though they may have been removed recently)
morgajelmy question is, why is it still asking about these hosts?
morgajelok... Think I just answered my own question.... the host DOES exist and the agent is running on it
voltermorgajel: These are the agents working in active mode, looking for configuration.
volterThey register by their given name in the agent configuration.
volterAnd if there is no match in the Zabbix database (actually the server configuration cache), this is what you get.
morgajelvolter: I just came to a similar conclusion- I think a colleague of mine mass installed the agent on a bunch of hosts, then turned it off on ones that had "issues". when we lost power in our DC and they all came on, the agents came back up and caused a ruckus
volterI'm not sure if this makes sense.
Richlvvolter, seems to make sense - they installed agent, stopped in some hosts. after a reboot, they started again :)
Richlvagents, hat is