Service under heavy load error

[enealDC]

At one of my installs, we've begun to see this error on both the primary and secondary system:

iSCSI Failover: Primary node: Serivice is under heavy load or the time values you have set are too low (interval 5400 ms). Please see Failover Configuration help for further details.

The system load on both nodes appear very low. We are currently running b4550.

I'm not sure how to interpret this error. What is the "service" under load?

[Gr-R]

The node is complaining about the heartbeat time intervals between nodes in failover.

see page 37 in the userguide about adjustments.

However, before making adjustments, you should update the nodes to the latest release.
Take care on the update process since its in failover configuration. Instructions in release notes.

[flix21]

Are the heartbeat options only available in v6? Is there a way to changed them in v5?

[Gr-R]

Are the heartbeat options only available in v6? Is there a way to changed them in v5?

Yes, see page 34 in the userguide.

[Al-S]

NOTE If you see the following error message: Service is under heavy load or the time values you have set are too low, you need to readjust the time values (they are dependent on your network connectivity and system load). When using low deadtine value, one of your nodes may wrongly assume the other one is dead, which will lead to a situation where both nodes are active (the so-called "split brain"). Please observe the following guidelines for deadtime value tuning:

- set the keepalive time to 1000 ms, it should be lower than warntime,
- set deadtime to 60 s,
- set warntime to half the deadtime value you wish to use,
- observe the messages in the error box (lower left-hand corner of the Web GUI) or your email, if you have set up email notification for failover. If there are no relevant warnings, your proposed deadtime is fine and you can skip to the next step. Otherwise, set warntime to the longest time interval between the hearbeat packets as indicated in the error box. If the errors persist, increase warntime further. Please note deadtime always needs to be greater than warntime,
- set deadtime to double the current warntime value,
- set warntime to slightly less than deadtime