Manual Chapter :
Troubleshooting Geo-Redundancy
Applies To:
Show Versions
BIG-IP LTM
- 17.1.0, 17.0.0, 16.1.1, 16.1.0, 16.0.1, 16.0.0, 15.0.1
Troubleshooting Geo-Redundancy
Reload to Recover
There are several reasons for noticeable discrepancies in persistence entries across
local and remote sites.
Option to force manual reload of all the Geo-Redundancy eligible sessiondb
entries from the remote site:
# geodb -r
Use this option with caution as it will override
all the local site entries with the reload entries from a remote site.
Undeploy and deploy iAppLX
If there are issues with configuration, refer to
REST APIs support for configuration and setup
commands
.Undeploy the iAppLX and deploy it again. This will result in installing the missing services
and updating the configuration correctly.
Reduce restnoded log level
The restnoded service runs in the background when iAppLX runs. If
/var/log/restnoded/restnoded*.log
takes a lot of space, it
is time to change the log level. Open the
/etc/bigstart/scripts/rest
noded file and update the
logLevel
to severe in the following line. Then restart the
restnoded service.# vim /etc/bigstart/scripts/restnoded # exec /usr/bin/f5-rest-node /usr/share/rest/node/src/restnode.js -p 8105 --logLevel severe -i ${LOG_FILE} -s none >> /var/tmp/${service}.out 2>&1 # bigstart restart restnoded
HA standby shows in sync, but iApp
does not show up on standby
Verify the detection of
both devices in the iApp framework device group. Both active and standby are in the
output:
restcurl
shared/resolver/device-groups/tm-shared-all-big-ips/devices
Both devices are in the ACTIVE state:
restcurl /shared/gossip
Active shows as active and standby shows as standby:
restcurl tm/shared/bigip-failover-state
If there is an issue with verifying the detection of both devices,
trigger the failover state again:
restcurl -X POST -d '{}'
tm/shared/bigip-failover-stat
iApp undeploy leaves the setup in a bad state (status RED in the
UI)
Before running the reinstallation process, follow these steps:
- 1. Clear the rest storage so that the iApp does not contain any old data:clear-rest-storage
- 2. Verify the/usr/bin/kafkadirectory is removed. This can happen if the previous iApp was not properly undeployed.
- 3. Get or correct failover states using the previous step (step 2).
Actives are not in alive and connected state
Reload
geodb
. If this does not work, try restarting
Kafka, ZooKeeper, and MirrorMaker1 a couple of times.// Reload geodb [root@bigip1:/S1-green-P::Active:In Sync] config # geodb -r // Restart Kafka, Zookeeper and Mirrormaker1 [root@bigip1:/S1-green-P::Active:In Sync] config # bigstart restart kafka zookeeper mirrormaker1
