Zabbix Server 3.4 - Late notification and data graph

[darklord20k]

Hi all,
Recently my zabbix is sending notification so late (as my attachment), not only email but also problems, triggers on web UI, graph data is not up to date too, but it still updating (but late), like I'm monitoring the past :lol: I don't know why, maybe syncer history process (always more than 75%)
I've tuning MySQL, zabbix_server.conf but still problem
Zabbix server was installed on Hyper V machine (2 cores processor, dynamic 2-8G RAM)
Please help! Thank you all!
my.cnf

!includedir /etc/mysql/conf.d/
!includedir /etc/mysql/mysql.conf.d/
[mysqld]
max_connections = 1000
query_cache_type = 1
innodb_flush_log_at_trx_commit = 2
optimizer_switch = index_condition_pushdown=off
max_heap_table_size = 256M
innodb_log_buffer_size = 256M
read_buffer_size = 256k
query_cache_limit = 1048576
read_rnd_buffer_size = 256k
innodb_buffer_pool_size = 2G
query_cache_size = 256M
tmp_table_size = 256M
slow-query-log
long_query_time = 20
slow_query_log_file = /var/log/mysql/slow-query.log
log_timestamps = SYSTEM
wait_timeout=300
interactive_timeout=60

zabbix_server.cnf

StartPollers=350
StartPollersUnreachable=40
StartTrappers=80
StartPingers=30
HousekeepingFrequency=8
MaxHousekeeperDelete=500
CacheSize=512M
StartDBSyncers=10
HistoryCacheSize=512M
HistoryIndexCacheSize=512M
TrendCacheSize=512M
ValueCacheSize=512M
Timeout=15
LogSlowQueries=1000

[kaspars.mednis]

Hi,

Which exact Zabbix Server version are you using ?

Regards,
Kaspars

[darklord20k]

Hi,

Which exact Zabbix Server version are you using ?

Regards,
Kaspars

Hi kaspars. I'm using ver 3.4.5
btw, I disabled housekeeper and re-tunned mysql server

max_connections = 1000

innodb_lru_scan_depth=256
innodb_flush_log_at_trx_commit = 2
innodb_log_buffer_size = 256M
innodb_log_file_size = 512M
innodb_buffer_pool_size = 4G
innodb_buffer_pool_instances=4

optimizer_switch = index_condition_pushdown=off

max_heap_table_size = 256M

join_buffer_size = 512k
read_buffer_size = 256k
read_rnd_buffer_size = 256k


tmp_table_size = 256M

query_cache_type = 0
query_cache_size = 0
query_cache_limit = 1048576

slow-query-log
long_query_time = 10
slow_query_log_file = /var/log/mysql/slow-query.log
log_timestamps = SYSTEM

wait_timeout=300
interactive_timeout=60

thread_cache_size = 10

after reboot ubuntu (long as hell, about more than 1 hours) zabbix is alive till now (~19 hours), include mysql restart 1 time ( 4 hours ago, i don't know why, log as bellow)
I'll update when zabbix suck again (hope no more )

2018-01-23T06:37:33.644930+07:00 0 [Note] InnoDB: page_cleaner: 1000ms intended loop took 4730ms. The settings might not be optimal. (flushed=306 and evicted=0, during the time.)
2018-01-23T06:37:41.224702+07:00 69638 [Note] Aborted connection 69638 to db: 'zabbix' user: 'zabbix' host: 'localhost' (Got timeout reading communication packets)
2018-01-23T06:37:46.416614+07:00 69961 [Note] Aborted connection 69961 to db: 'zabbix' user: 'zabbix' host: 'localhost' (Got timeout reading communication packets)
2018-01-23T06:37:56.197252+07:00 69582 [Note] Aborted connection 69582 to db: 'zabbix' user: 'zabbix' host: 'localhost' (Got timeout reading communication packets)
2018-01-23T06:38:00.836934+07:00 69990 [Note] Aborted connection 69990 to db: 'zabbix' user: 'zabbix' host: 'localhost' (Got timeout reading communication packets)
2018-01-23T06:38:14.185443+07:00 70010 [Note] Aborted connection 70010 to db: 'zabbix' user: 'zabbix' host: 'localhost' (Got timeout reading communication packets)
2018-01-23T06:38:17.245433+07:00 69669 [Note] Aborted connection 69669 to db: 'zabbix' user: 'zabbix' host: 'localhost' (Got timeout reading communication packets)
2018-01-23T06:38:27.897516+07:00 70018 [Note] Aborted connection 70018 to db: 'zabbix' user: 'zabbix' host: 'localhost' (Got timeout reading communication packets)
2018-01-23T06:38:32.249859+07:00 69709 [Note] Aborted connection 69709 to db: 'zabbix' user: 'zabbix' host: 'localhost' (Got timeout reading communication packets)
2018-01-23T06:38:39.972093+07:00 70031 [Note] Aborted connection 70031 to db: 'zabbix' user: 'zabbix' host: 'localhost' (Got timeout reading communication packets)
2018-01-23T06:38:52.043514+07:00 70038 [Note] Aborted connection 70038 to db: 'zabbix' user: 'zabbix' host: 'localhost' (Got timeout reading communication packets)
2018-01-23T06:39:04.982270+07:00 70068 [Note] Aborted connection 70068 to db: 'zabbix' user: 'zabbix' host: 'localhost' (Got timeout reading communication packets)
2018-01-23T06:39:15.945638+07:00 0 [Note] InnoDB: page_cleaner: 1000ms intended loop took 7220ms. The settings might not be optimal. (flushed=251 and evicted=0, during the time.)
2018-01-23T06:39:21.196227+07:00 70090 [Note] Aborted connection 70090 to db: 'zabbix' user: 'zabbix' host: 'localhost' (Got timeout reading communication packets)
2018-01-23T06:39:27.861554+07:00 70097 [Note] Aborted connection 70097 to db: 'zabbix' user: 'zabbix' host: 'localhost' (Got timeout reading communication packets)
2018-01-23T06:39:39.337540+07:00 70110 [Note] Aborted connection 70110 to db: 'zabbix' user: 'zabbix' host: 'localhost' (Got timeout reading communication packets)
2018-01-23T06:39:49.536169+07:00 0 [Note] Giving 61 client threads a chance to die gracefully
2018-01-23T06:39:49.536290+07:00 0 [Note] Shutting down slave threads
2018-01-23T06:39:51.542100+07:00 0 [Note] Forcefully disconnecting 55 remaining clients
2018-01-23T06:39:51.542161+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70382 user: 'zabbix'
2018-01-23T06:39:51.542203+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70145 user: 'zabbix'
2018-01-23T06:39:51.542237+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70434 user: 'zabbix'
2018-01-23T06:39:51.542263+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70217 user: 'zabbix'
2018-01-23T06:39:51.542288+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 67 user: 'zabbix'
2018-01-23T06:39:51.542313+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70347 user: 'zabbix'
2018-01-23T06:39:51.542344+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 3 user: 'zabbix'
2018-01-23T06:39:51.542370+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70428 user: 'zabbix'
2018-01-23T06:39:51.542394+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70386 user: 'zabbix'
2018-01-23T06:39:51.542425+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70262 user: 'zabbix'
2018-01-23T06:39:51.542451+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70192 user: 'zabbix'
2018-01-23T06:39:51.542476+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 5 user: 'zabbix'
2018-01-23T06:39:51.542521+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 22 user: 'zabbix'
2018-01-23T06:39:51.542547+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70289 user: 'zabbix'
2018-01-23T06:39:51.542571+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70177 user: 'zabbix'
2018-01-23T06:39:51.542595+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70335 user: 'zabbix'
2018-01-23T06:39:51.542626+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70352 user: 'zabbix'
2018-01-23T06:39:51.542651+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70298 user: 'zabbix'
2018-01-23T06:39:51.542675+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70500 user: 'zabbix'
2018-01-23T06:39:51.542706+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70366 user: 'zabbix'
2018-01-23T06:39:51.542737+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70341 user: 'zabbix'
2018-01-23T06:39:51.542761+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70342 user: 'zabbix'
2018-01-23T06:39:51.542785+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70044 user: 'zabbix'
2018-01-23T06:39:51.542815+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70395 user: 'zabbix'
2018-01-23T06:39:51.542846+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70326 user: 'zabbix'
2018-01-23T06:39:51.542870+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70450 user: 'zabbix'
2018-01-23T06:39:51.542900+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 37 user: 'zabbix'
2018-01-23T06:39:51.542925+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70116 user: 'zabbix'
2018-01-23T06:39:51.542955+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70163 user: 'zabbix'
2018-01-23T06:39:51.542985+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 47 user: 'zabbix'
2018-01-23T06:39:51.543015+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70471 user: 'zabbix'
2018-01-23T06:39:51.543039+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 13 user: 'zabbix'
2018-01-23T06:39:51.543069+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70418 user: 'zabbix'
2018-01-23T06:39:51.543100+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70072 user: 'zabbix'
2018-01-23T06:39:51.543124+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70094 user: 'zabbix'
2018-01-23T06:39:51.543148+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70492 user: 'zabbix'
2018-01-23T06:39:51.543179+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70497 user: 'zabbix'
2018-01-23T06:39:51.543203+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 64 user: 'zabbix'
2018-01-23T06:39:51.543226+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70236 user: 'zabbix'
2018-01-23T06:39:51.543256+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70498 user: 'zabbix'
2018-01-23T06:39:51.543287+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70422 user: 'zabbix'
2018-01-23T06:39:51.543317+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 55 user: 'zabbix'
2018-01-23T06:39:51.543347+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70416 user: 'zabbix'
2018-01-23T06:39:51.543377+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 69 user: 'zabbix'
2018-01-23T06:39:51.543402+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 38 user: 'zabbix'
2018-01-23T06:39:51.543432+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70186 user: 'zabbix'
2018-01-23T06:39:51.543463+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70465 user: 'zabbix'
2018-01-23T06:39:51.543493+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70243 user: 'zabbix'
2018-01-23T06:39:51.543518+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70036 user: 'zabbix'
2018-01-23T06:39:51.543557+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70306 user: 'zabbix'
2018-01-23T06:39:51.543588+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70266 user: 'zabbix'
2018-01-23T06:39:51.543612+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70424 user: 'zabbix'
2018-01-23T06:39:51.543643+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70258 user: 'zabbix'
2018-01-23T06:39:51.543673+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70383 user: 'zabbix'
2018-01-23T06:39:51.544480+07:00 0 [Warning] /usr/sbin/mysqld: Forcing close of thread 70339 user: 'zabbix'
2018-01-23T06:39:51.545645+07:00 0 [Note] Event Scheduler: Purging the queue. 0 events
2018-01-23T06:39:51.989348+07:00 0 [Note] Binlog end
2018-01-23T06:39:52.315977+07:00 0 [Note] Shutting down plugin 'ngram'
2018-01-23T06:39:52.316075+07:00 0 [Note] Shutting down plugin 'BLACKHOLE'
2018-01-23T06:39:52.316115+07:00 0 [Note] Shutting down plugin 'partition'
2018-01-23T06:39:52.316141+07:00 0 [Note] Shutting down plugin 'ARCHIVE'
2018-01-23T06:39:52.316200+07:00 0 [Note] Shutting down plugin 'MRG_MYISAM'
2018-01-23T06:39:52.325035+07:00 0 [Note] Shutting down plugin 'INNODB_SYS_VIRTUAL'
2018-01-23T06:39:52.333259+07:00 0 [Note] Shutting down plugin 'INNODB_SYS_DATAFILES'
2018-01-23T06:39:52.333304+07:00 0 [Note] Shutting down plugin 'INNODB_SYS_TABLESPACES'
2018-01-23T06:39:52.333328+07:00 0 [Note] Shutting down plugin 'INNODB_SYS_FOREIGN_COLS'
2018-01-23T06:39:52.333349+07:00 0 [Note] Shutting down plugin 'INNODB_SYS_FOREIGN'
2018-01-23T06:39:52.333405+07:00 0 [Note] Shutting down plugin 'INNODB_SYS_FIELDS'
2018-01-23T06:39:52.333432+07:00 0 [Note] Shutting down plugin 'INNODB_SYS_COLUMNS'
2018-01-23T06:39:52.333453+07:00 0 [Note] Shutting down plugin 'INNODB_SYS_INDEXES'
2018-01-23T06:39:52.333474+07:00 0 [Note] Shutting down plugin 'INNODB_SYS_TABLESTATS'
2018-01-23T06:39:52.333496+07:00 0 [Note] Shutting down plugin 'INNODB_SYS_TABLES'
2018-01-23T06:39:52.333518+07:00 0 [Note] Shutting down plugin 'INNODB_FT_INDEX_TABLE'
2018-01-23T06:39:52.333541+07:00 0 [Note] Shutting down plugin 'INNODB_FT_INDEX_CACHE'
2018-01-23T06:39:52.333562+07:00 0 [Note] Shutting down plugin 'INNODB_FT_CONFIG'
2018-01-23T06:39:52.333586+07:00 0 [Note] Shutting down plugin 'INNODB_FT_BEING_DELETED'
2018-01-23T06:39:52.333608+07:00 0 [Note] Shutting down plugin 'INNODB_FT_DELETED'
2018-01-23T06:39:52.333629+07:00 0 [Note] Shutting down plugin 'INNODB_FT_DEFAULT_STOPWORD'
2018-01-23T06:39:52.333649+07:00 0 [Note] Shutting down plugin 'INNODB_METRICS'
2018-01-23T06:39:52.333669+07:00 0 [Note] Shutting down plugin 'INNODB_TEMP_TABLE_INFO'
2018-01-23T06:39:52.333689+07:00 0 [Note] Shutting down plugin 'INNODB_BUFFER_POOL_STATS'
2018-01-23T06:39:52.333711+07:00 0 [Note] Shutting down plugin 'INNODB_BUFFER_PAGE_LRU'
2018-01-23T06:39:52.333734+07:00 0 [Note] Shutting down plugin 'INNODB_BUFFER_PAGE'
2018-01-23T06:39:52.333756+07:00 0 [Note] Shutting down plugin 'INNODB_CMP_PER_INDEX_RESET'
2018-01-23T06:39:52.333775+07:00 0 [Note] Shutting down plugin 'INNODB_CMP_PER_INDEX'
2018-01-23T06:39:52.333794+07:00 0 [Note] Shutting down plugin 'INNODB_CMPMEM_RESET'
2018-01-23T06:39:52.333814+07:00 0 [Note] Shutting down plugin 'INNODB_CMPMEM'
2018-01-23T06:39:52.333833+07:00 0 [Note] Shutting down plugin 'INNODB_CMP_RESET'
2018-01-23T06:39:52.333853+07:00 0 [Note] Shutting down plugin 'INNODB_CMP'
2018-01-23T06:39:52.333874+07:00 0 [Note] Shutting down plugin 'INNODB_LOCK_WAITS'
2018-01-23T06:39:52.333895+07:00 0 [Note] Shutting down plugin 'INNODB_LOCKS'
2018-01-23T06:39:52.333915+07:00 0 [Note] Shutting down plugin 'INNODB_TRX'
2018-01-23T06:39:52.333935+07:00 0 [Note] Shutting down plugin 'InnoDB'
2018-01-23T06:39:52.347505+07:00 0 [Note] InnoDB: FTS optimize thread exiting.
2018-01-23T06:39:52.347781+07:00 0 [Note] InnoDB: Starting shutdown...
2018-01-23T06:39:52.456313+07:00 0 [Note] InnoDB: Dumping buffer pool(s) to /var/lib/mysql/ib_buffer_pool
2018-01-23T06:39:52.505484+07:00 0 [Note] InnoDB: Buffer pool(s) dump completed at 180123 6:39:52

[darklord20k]

After 2 days uptime, this morning, my zabbix get problems again, some graph on some host don't display data and notification right time (still displaying, but very slow and out-of-date), include zabbix server
Maybe MySQL problem? how can i fix it? May I manually execute Housekeeper?
Please help me I wan't to monitor realtime, not the past

[Clontarf[X]]

Hi There,

We have experienced exactly the same problem with our instance (reasonably large, 10k hosts, 1.7 billion items). Can you please do me a favour and tell me what the context switching statistics are like for your Zabbix Application server and your Database server?

What we see is that we have more than enough resources allocated to the Application server (32 cores, 256GB ram), however our context switching (due to number of threads for pollers/trappers/e.t.c) is extremely high.

When we see a large jump in context switching, we see a corresponding introduction of value delays (the issue you've described). We're not yet sure which comes first, the chicken or the egg.

Worth noting, we are running Zabbix 3.2, so it's not a version specific issue.

[darklord20k]

Hi There,

We have experienced exactly the same problem with our instance (reasonably large). Can you please do me a favour and tell me what the context switching statistics are like for your Zabbix Application server and your Database server?

What we see is that we have more than enough resources allocated to the Application server (32 cores, 256GB ram), however our context switching (due to number of threads for pollers/trappers/e.t.c) is extremely high.

When we see a large jump in context switching, we see a corresponding introduction of value delays (the issue you've described). We're not yet sure which comes first, the chicken or the egg.

Thank for your quick reply,
I'm running both zabbix server and MySQL on same host (Ubunt 16.04 on Hyper-V, 2 cores, 2.53 GHz, maximum 8GB RAM, HDD raid 5)
I'm new to zabbix and ubuntu, feel free to ask me if you need more information (and tell me how to get them, if you don't mind )

[Clontarf[X]]

Thank for your quick reply,
I'm running both zabbix server and MySQL on same host (Ubunt 16.04 on Hyper-V, 2 cores, 2.53 GHz, maximum 8GB RAM, HDD raid 5)
I'm new to zabbix and ubuntu, feel free to ask me if you need more information (and tell me how to get them, if you don't mind )

Ubuntu Context Switching
Item: system.cpu.switches

Assign that to your Zabbix/DB host and watch it for a few days, and then see what the values are at the time of the delays beginning.

[darklord20k]

Ubuntu Context Switching
Item: system.cpu.switches

Assign that to your Zabbix/DB host and watch it for a few days, and then see what the values are at the time of the delays beginning.

Look like value is same as when not delay

[Clontarf[X]]

Perhaps context switching is not the common problem here then, interesting.

What is interesting though is you do run a high number of poller processes (we run 450 for far more hosts than it looks like you have).

### Option: StartPollers
# StartPollers=5
StartPollers=450

### Option: StartIPMIPollers
# StartIPMIPollers=0
StartIPMIPollers=5

### Option: StartPollersUnreachable
# StartPollersUnreachable=1
StartPollersUnreachable=75

### Option: StartHTTPPollers
# StartHTTPPollers=1
StartHTTPPollers=5

### Option: StartJavaPollers
# StartJavaPollers=0
StartJavaPollers=0

### Option: StartProxyPollers
# StartProxyPollers=1
StartProxyPollers=5

Our latest delayed value period is shown in orange, so this supports your data that a change in context switching isn't necessarily the problem.

Zabbix DB ---------------------------------- Zabbix Server

So so far in common, we only have the number of poller processes.

The last time we had delayed values, we added 50+ new items to a template which was applied to hundreds of hosts. It seems as soon as this operation started, values were delayed.

[Clontarf[X]]

Actually, can you do me another favour, can you show me what your History buffers look like during delayed value periods? In our circumstance, we see that the History Buffer grows while the values are delayed, and returns to normal values when the values have caught up. See below (again, delayed value periods in orange).

[darklord20k]

Actually, can you do me another favour, can you show me what your History buffers look like during delayed value periods? In our circumstance, we see that the History Buffer grows while the values are delayed, and returns to normal values when the values have caught up. See below (again, delayed value periods in orange).

many thanks for your passion,
I don't find extract items like your History Buffer, can you tell me extract it's key.
BTW, more start poller is better, isn't it?

[darklord20k]

Perhaps context switching is not the common problem here then, interesting.

What is interesting though is you do run a high number of poller processes (we run 450 for far more hosts than it looks like you have).

### Option: StartPollers
# StartPollers=5
StartPollers=450

### Option: StartIPMIPollers
# StartIPMIPollers=0
StartIPMIPollers=5

### Option: StartPollersUnreachable
# StartPollersUnreachable=1
StartPollersUnreachable=75

### Option: StartHTTPPollers
# StartHTTPPollers=1
StartHTTPPollers=5

### Option: StartJavaPollers
# StartJavaPollers=0
StartJavaPollers=0

### Option: StartProxyPollers
# StartProxyPollers=1
StartProxyPollers=5

Our latest delayed value period is shown in orange, so this supports your data that a change in context switching isn't necessarily the problem.

Zabbix DB ---------------------------------- Zabbix Server

So so far in common, we only have the number of poller processes.

The last time we had delayed values, we added 50+ new items to a template which was applied to hundreds of hosts. It seems as soon as this operation started, values were delayed.

I've increased start poller to 300, reboot server and data still delay, I think problems is somewhere beetween data buffer and MySQL.
Devices send snmp to buffer success, but data transfer from buffer to MySQL slow then zabbix grab outdated data. But I don't know how to improve it

[kaspars.mednis]

Hi,

Late data graph can be caused by history syncer issues. Zabbix Server 3.4.5 may be affected by

https://support.zabbix.com/browse/ZBX-13343

in the graph named Zabbix internal process busy pay attention to history syncer usage , if it is 100% busy, that's the cause of the problem.

Possible workaround - you can try downgrade to 3.4.4 , or wait for 3.4.7

Regards,
Kaspars

[darklord20k]

Hi,

Late data graph can be caused by history syncer issues. Zabbix Server 3.4.5 may be affected by

https://support.zabbix.com/browse/ZBX-13343

in the graph named Zabbix internal process busy pay attention to history syncer usage , if it is 100% busy, that's the cause of the problem.

Possible workaround - you can try downgrade to 3.4.4 , or wait for 3.4.7

Regards,
Kaspars

oh s*** bad news for 2018

[darklord20k]

Update: I rebuild server (same hyper-V host, same setting, but on debian 9)
running for 1 weeks sounds good.