Legacy Graphite¶
Imported from Confluence
Content may be outdated. Verify before following any procedures. View original | Last updated: January 2026
Known Issues¶
No Data alerts¶
We are getting No Data alerts when graphite is down . It's still running on VM - graphite-core.vm-offerwall-prod-useast1.fyber.com.
In case you are not able to ssh we need to do hard reset from the console. In such case temporary folders will be removed from OS.
Info
For SSH access, ask the DevOps BLN team
Graphite store PID number of the process in the file in one of temporary folders. In order to fix.
- Check that file exists, if not - create folder and change owner.
cat /var/run/carbon/carbon1.pid
cat /var/run/carbon/carbon2.pid
mkdir /var/run/carbon/
chown -R graphite:graphite /var/run/carbon
- Start graphite process
- Check logs
Graphite host out of disk space¶
SSH to the instance and clean some space
If the above doesn't help, it might be because Graphite machine has run out of space. Check under /storage/statsite to get some clue about the files that are getting accumulated. For example:
cd /storage/statsite/counts/k8/engine
find . -maxdepth 1 -type d -name 'engine-*' -newermt "2018-01-01" ! -newermt "2025-05-20" -exec rm -r {} +
or:
# Confirm first that blacklisted IPs are okay to remove
cd /storage/statsite/counts/sws/rails/vm-ofw-core-offerwall-prod-useast1-0/blacklisted_ips
rm -r *
cd /storage/statsite/counts/sws/rails/vm-ofw-core-offerwall-prod-useast1-1/blacklisted_ips
rm -r *
cd /storage/statsite/counts/sws/rails/vm-ofw-core-offerwall-prod-useast1-2/blacklisted_ips/
rm -r *
...
# Continue with the rest of machines
HostOutOfMemory¶
Alert in #bln-escalation Slack channel:
image-2025-11-24_18-3-47.png¶
Probably you need to roload or restart httpd service on VM.
List Failed services and reset