-
|
Hello I am not sure it is a bug or a misconfiguration, what I did
for what I saw after the upgrade
Before the upgrade I can list few emails and when I send emails to piler I can see the counter increasing in the admin page. moreover I can see in the postfix log the messages about sending the email to piler I took the time to diff all the configuration files in /etc/piler with a direct install and a piler upgrade and I have the same content. What is puzzling me is that the same configuration but as a direct install is workable, so my configuration should be good |
Beta Was this translation helpful? Give feedback.
Replies: 16 comments 4 replies
-
|
hum so piler after the upgrade cannot connect to mysl :/ |
Beta Was this translation helpful? Give feedback.
-
|
hum I suspect really a bug due to #278 and another one, because after the upgrade I have also what I did
I check the permissions I set the permissions to piler:piler on store with -R and everything is up and workable b ut I am not sure that I have the old index, I need maybe to wait the cron |
Beta Was this translation helpful? Give feedback.
-
|
relative to the error on I have I think we have a bug somewhere with the upgrade @jsuto |
Beta Was this translation helpful? Give feedback.
-
|
I think the problem is that the ubuntu 24.04 docker image ships the ubuntu user who's uid is the same as it was piler's in the the sutoj/piler:1.4.4 image that was built on ubuntu 22.04. The fix could be either running chown piler:piler for these files and directories, or fix /etc/piler/piler.conf username variable to run as user ubuntu. Not sure which is easier in your environment. I'm biased to keep piler as the running user. |
Beta Was this translation helpful? Give feedback.
-
|
Or I may run |
Beta Was this translation helpful? Give feedback.
-
|
Hello I got with my environment the way to trigger actions during upgrade, before and after the restart of the service, so I did after the service restart a chown myself, unfortunately it is not enough and piler cannot connect to mysql, until I restart piler a second time after the upgrade :p |
Beta Was this translation helpful? Give feedback.
-
|
I think the ownership of this ubuntu should be fixed @jsuto, maybe by changing the UID/GID of piler itself, or by removing the ubuntu user but probably piler needs to become a sudo user itself The upgrade is not easy I understand, but it is something we have to do/follow |
Beta Was this translation helpful? Give feedback.
-
|
I've just created a fixed docker image, sutoj/piler:1.4.7b that has piler user with uid/gid: 1000/1000, just as with the previous release based on ubuntu 22.04. Let me know how it goes. If the upgrade is successful, then I'll release 1.4.8. |
Beta Was this translation helpful? Give feedback.
-
|
at the first tests it seems OK, thank a lot, can you consider also to move the tempates you are using outside of /etc/piler that is a recommended mounting volume. I mean all *.dist files since it is a mounted volume, we keep the previous version and when you change for example the version of php8.1 to 8.3 in piler-nginx.conf.dist we continue to use the version 81, the container upgrade does not change the file piler-nginx.conf.dist These templates files should not be in a mounted volume I continue my tests |
Beta Was this translation helpful? Give feedback.
-
|
ok now the upgrade is not more affected by the ownership issue, however all the previous searches that were available in the auditor view are removed. we have a Is that an expected behavior @jsuto after the upgrade or I missed and I need to fix something |
Beta Was this translation helpful? Give feedback.
-
|
Can you please elaborate the "we have no search view" issue? Losing the index data after the upgrade is not expected. |
Beta Was this translation helpful? Give feedback.
-
|
in the piler container we have this ID
in the manticore container we have this ID
So yes definitely we have some conflicts, this explain why I need to use |
Beta Was this translation helpful? Give feedback.
-
|
noob question @jsuto this is what I found in the old container 1.4.4, I think I use This is what I find in the version 1.4.7b with RT=1 Do you think it is compatible, since I do not use the same RT options ????? |
Beta Was this translation helpful? Give feedback.
-
|
after the upgrade from 1.4.4 to 1.4.7b I find well the files to /var/lib/manticore with the good ownership |
Beta Was this translation helpful? Give feedback.
-
|
I use the default configuration |
Beta Was this translation helpful? Give feedback.
-
|
for the record, I fixed my issue by starting again to index the emails to piler after the change of the manticore to a standalone container
|
Beta Was this translation helpful? Give feedback.
for the record, I fixed my issue by starting again to index the emails to piler after the change of the manticore to a standalone container
podman exec -w /var/piler/imap piler-app reindex -a -pNethServer/ns8-piler#37