This website uses cookies. By clicking Accept, you consent to the use of cookies. Click Here to learn more about how we use cookies.
Accept
Reject
Help
cancel
Turn on suggestions
Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.
Showing results for 
Search instead for 
Did you mean: 

Hi, We are using HM version 19.5.1.7-NGVA on prem. Since a week, we have the problem that the System Service “Elasticsearch” is not running.

Hi, We are using HM version 19.5.1.7-NGVA on prem. Since a week, we have the problem that the System Service “Elasticsearch” is not running.

it_admin
New Contributor

‎07-29-2019 10:18 AM

Hi, We are using HM version 19.5.1.7-NGVA on prem. Since a week, we have the problem that the System Service “Elasticsearch” is not running.
0 Kudos
9 REPLIES 9

rpjunior
New Contributor

‎08-13-2019 01:34 PM

I have the same problem.

After upgrade to 19.5 version, elasticsearch has been problem.

I reboot the system and o system works well for a short period until elasticsearch crash it again.

0 Kudos

weekdaysailor
Contributor

‎07-30-2019 05:35 PM

Another very common culprit w daemons dying is the logging partition (e.g /var/log ) filling up. When they cannot write to their log, they usually terminate. You would probably need to engage with Aerohive support to address this as it probably requires shell access.

 

Did you give it the Aerohive recommended disk allocation? Is it static or dynamic?

0 Kudos

it_admin
New Contributor

‎07-30-2019 07:37 AM

Hi Keith,

yes it was working fine. There are no changes on the vmware Host.

There are a couple of machines running, but they do not need very much CPU time most of the time they are idle.

The only time when the CPU is busy is when the Hivemanager starts.

Regards

Mike

0 Kudos

weekdaysailor
Contributor

‎07-29-2019 03:42 PM

 

Since you say this is going on for a week (and it was working fine before?), this symptom is consistent with an over-subscribed virtual host. So even though you've given your VM "4 CPU's" - that's not really how it works. You're really just giving it Ghz (clock cycles) that are shared with other virtual hosts on that (bare metal) machine. If there are other VMs on the host that are busy, they can consume "unfairly" depending on the configuration. So I would look to the virtual host performance data and see if there is evidence of over-subscription (and most can also allow you to prioritize the allocations) - e.g. if the host CPU cores are hitting 100% on a regular basis..it's over-subscribed. (memory works a little different but can have the same effect)

 

-K

 

0 Kudos

it_admin
New Contributor

‎07-29-2019 10:19 AM

If we restart all services, it takes up to 30 minutes and all Services are up and running again.

After a few minutes, the service stops.

We have that problem with an older Version of the Hivemanger before.

That issue was fixed with version 12.8.3.1.

We also tried to reboot the machine but that has the same effect like the restart.

The virtual machine is running with 4 CPU and 24GB RAM. We are managing 100 Aerohive devices.

Any idears?

 

0 Kudos
GTM-P2G8KFN