ok, one step forward, two steps backwards now
Two more commands for you to try:
package-cleanup --cleandupes --skip-broken
yum update
ok, one step forward, two steps backwards now
Two more commands for you to try:
package-cleanup --cleandupes --skip-broken
yum update
Looks like it deleted 195 items and updated/cleaned/verified 154 items. Looks like I cant access the Web GUI anymore but I can still via SSH.
ok, output would be helpful. Whatās the output if you run yum update
one more time?
when I ran package-cleanup --cleandupes --skip-broken
It deleted 195 items but I didnt save the logs from it but here are the logs for yum update the first time I ran it and the second time I ran it.
EDIT: I rean yum info rockstor
and it out came back with this. So In theory I should be on build 11 right?
Great. Itās looking promising. I think every package is now updated and yum is happy. itās just a matter of starting/restarting rockstor services. You can do that with
systemctl restart rockstor-pre
and wait until it shows āactiveā in systemctl status -l rockstor-pre
systemctl restart rockstor
and wait till active like abovesystemctl restart rockstor-bootstrap
and wait like aboveā¦Please do report back the output of status commands for sanity check. If all goes well, web-ui should be back online.
I got the follow error when I tried to run systemctl restart rockstor-pre
Job for rockstor-pre.service failed because the control process exited with error code. See "systemctl status rockstor-pre.service" and "journalctl -xe" for details.
ok, we are not completely out of the woods yet. Please share the output of systemctl status -l rockstor-pre
and journalctl -xe
commands.
systemctl status -l rockstor-pre and journalctl -xe
is 1241 lines long. Pastebin was unhappy with it being that long. Is there anything specific youāre looking for or an easy way to save it as a .txt and just retrieve it?
the key line was included in the output
Jan 28 14:20:49 nas systemd[8786]: Failed at step EXEC spawning /opt/rockstor/bin/initrock: No such file or directory
This is pretty bizarre. /opt/rockstor is where all the files of the rockstor package live. Did yum completely nuke that directory?
This doesnāt make sense because yum info shows that 3.8-11 is installed. After verifying the contents of /opt/rockstor, you could try yum update rockstor
and see if that pulls the files again?
I looked in /opt/rockstor and the only folders were certs src static var
and the update reports back
`[root@nas rockstor]# yum update rockstor
Loaded plugins: changelog, fastestmirror
Rockstor-Stable | 2.9 kB 00:00
rockstor | 2.9 kB 00:00
Loading mirror speeds from cached hostfile
bummer. that does not look good. Rockstor got uninstalled some how. You can try yum install rockstor
and that should install 3.8-11. But I am afraid the application db may not be reusable. But give it a shot and see what happens with rockstor-pre after that. If rockstor-pre start/restarts fine, itās good(unlikely, though). If not, youāll have to re-install LMK.
Looks like you might be right. I did yum install rockstor and it seemed to work, but after I rebooted the machine I got tons of could not boot errors. If I go and reinstall am I going to need a new key? Despite the issues I had I canāt thank you enough for trying to get me sorted out. I really appreciate it. Support like this is why I donāt mind paying the extra dollars for it.
Itās a disappointing end to our troubleshooting exercise but I am pleased with your attitude and really appreciate all the feedback. Here are the top takeaways I am going to incorporate into future development
After re-install, you will need a new key. Just e-mail your appliance ID to support@rockstor.com and weāll send you one.
No worries. I wish I could have helped more. At least we both learning something in the process so its not all in vain. I understand that itās still in a development state and issues come with it.
Good day @suman.
Just got this on my fresh install (well fresh from couple weeks) as an alert. Following the instructions got the āproblem solvedā.
Have the team made any progress on solving this without operator intervention? Iām afraid some users my break things going to the shell
Best regards,
Al