Dashboard graphs not available
Pages: 1 2
admin
2,930 Posts
September 12, 2017, 3:43 pmQuote from admin on September 12, 2017, 3:43 pmYes it will help, suggest the following sequence
systemctl stop petasan-mount-sharedfs
umount -l /opt/petasan/config/shared/
systemctl stop glusterfs-server
killall glusterfsd
killall glusterfs
systemctl start glusterfs-server
systemctl start petasan-mount-sharedfs
Yes it will help, suggest the following sequence
systemctl stop petasan-mount-sharedfs
umount -l /opt/petasan/config/shared/
systemctl stop glusterfs-server
killall glusterfsd
killall glusterfs
systemctl start glusterfs-server
systemctl start petasan-mount-sharedfs
Last edited on September 12, 2017, 3:43 pm by admin · #11
admin
2,930 Posts
September 13, 2017, 12:14 pmQuote from admin on September 13, 2017, 12:14 pmWe fixed an issue that may be related to the problem: The dashboard used to send ajax requests to refresh ceph status periodically, but in case there was a long response time for any reason (network or system load) the requests will keep being send periodically even if older ones did not complete. So there is potential for these requests to pile up and maybe cause memory and connection load on the monitors. Maybe it relates to your case.
The attached fix dis-ables any further status requests until previous one completes.
https://drive.google.com/file/d/0B7VNYCjYBY2yeGJ2UjQxZTJMYWs/view?usp=sharing
There is a fix for both v 1.31 and 1.40
We fixed an issue that may be related to the problem: The dashboard used to send ajax requests to refresh ceph status periodically, but in case there was a long response time for any reason (network or system load) the requests will keep being send periodically even if older ones did not complete. So there is potential for these requests to pile up and maybe cause memory and connection load on the monitors. Maybe it relates to your case.
The attached fix dis-ables any further status requests until previous one completes.
https://drive.google.com/file/d/0B7VNYCjYBY2yeGJ2UjQxZTJMYWs/view?usp=sharing
There is a fix for both v 1.31 and 1.40
Last edited on September 13, 2017, 12:23 pm by admin · #12
Pages: 1 2
Dashboard graphs not available
admin
2,930 Posts
Quote from admin on September 12, 2017, 3:43 pmYes it will help, suggest the following sequence
systemctl stop petasan-mount-sharedfs
umount -l /opt/petasan/config/shared/
systemctl stop glusterfs-server
killall glusterfsd
killall glusterfs
systemctl start glusterfs-server
systemctl start petasan-mount-sharedfs
Yes it will help, suggest the following sequence
systemctl stop petasan-mount-sharedfs
umount -l /opt/petasan/config/shared/
systemctl stop glusterfs-server
killall glusterfsd
killall glusterfs
systemctl start glusterfs-server
systemctl start petasan-mount-sharedfs
admin
2,930 Posts
Quote from admin on September 13, 2017, 12:14 pmWe fixed an issue that may be related to the problem: The dashboard used to send ajax requests to refresh ceph status periodically, but in case there was a long response time for any reason (network or system load) the requests will keep being send periodically even if older ones did not complete. So there is potential for these requests to pile up and maybe cause memory and connection load on the monitors. Maybe it relates to your case.
The attached fix dis-ables any further status requests until previous one completes.
https://drive.google.com/file/d/0B7VNYCjYBY2yeGJ2UjQxZTJMYWs/view?usp=sharing
There is a fix for both v 1.31 and 1.40
We fixed an issue that may be related to the problem: The dashboard used to send ajax requests to refresh ceph status periodically, but in case there was a long response time for any reason (network or system load) the requests will keep being send periodically even if older ones did not complete. So there is potential for these requests to pile up and maybe cause memory and connection load on the monitors. Maybe it relates to your case.
The attached fix dis-ables any further status requests until previous one completes.
https://drive.google.com/file/d/0B7VNYCjYBY2yeGJ2UjQxZTJMYWs/view?usp=sharing
There is a fix for both v 1.31 and 1.40