Search results

This is crayz. Brandy new hardware and on 3rd day, I am getting this issue.

Just to make sure, this is not about ksm sharing ? So can we say, if windows does not leave much free ram, instead use for caching then it thinks it is better to do so to gaster things...

Should I disable or clean occasionally (if possible), that caching ? Because it makes me think there is no enough ram on vm. Is this something related to ksm or balooning setting? Or only with windows?

So you mean, I must also count that "Cached" amount to total ram in use?

When I check memory usage in windows, total memory in use does not sound correct. Please see screen shot, that vm has total 20 GB Ram but windows show %85 in use which means 17 GB but if you look at processes it is not in total 17 GB. Maybe around 8-9 GB. Using zfs..

I am having same problem with all brand new hardwares for 3-4 proxmox nodes. So I really do not think it is realted to hardware. Luckly it happens only 1-2 times in 2-3 months.. Only thing in logs is: -- Reboot -- And yes there are many threads in forum.

I think I also do not need to add channel or even maybe custom ruleset. Thank you.

- What config would you use for this? I can set high scores in sa rules and let pmg mark it as spam or even block mail. I wanted add spamassassin.heinlein-support.de channel (not sure if it good idea yet, here it was mentioned, you can also share your ideas about it) but I could not. Because I...

Thank you but I can also block mails with SA rules. So should I use actions for sure actions like blocking, and use SA rules for small header score adds. ? Also, should I think of adding additional SA channels / custom rules ? Or actions frum gui is all ok for.?

What is the difference of using builtin rulesets from gui or spamassian rules custom /etc/mail/spamassassin/custom.cf ? Performance / usage / easy to manage?

Just wanted to update here about that I have been running PMG for couple days and till now I am happy with it. So need some attention for installing and adjusting detailed configuration. So I highly can recomment it for anyone still trying to decide about it. :)

If it was a temprature issue, would I see it in logs?

Are you still having issue ? I just checked my older notes, same thing happened one month ago.

So that mean no error is logged. If it was an sudden electricy reset, would logs be like this?

I do not think it has frozen. Just rebooted wht only -- Reboot -- in logs. I asked guys in DC if someone has rebotted it but it is not possible. So what does that "-- Reboot --" text in log mean I wonder. Hardware is new.

I loose connection in server and have guys in DC to rebooted server. Then when I check logs only thing I see at connection lost time: Oct 31 12:00:20 s8 CRON[911689]: pam_unix(cron:session): session closed for user root -- Reboot -- Oct 31 12:22:11 s8 kernel: Linux version 5.15.64-1-pve...

Not sure about kernel update. But I can say I updated kernel about 2 weeks ago. There are 15 GB free ram of 128 GB total. But there are some backup zfs snapshot from other nodes. They do not need ram when not in use right? I will consider limiting disk. But I saw many similar threads in forum...

No specific task. But one vm is doing high disk io. (12 TB Disk read in 8 days with nvme speed.) I set disk type to Async IO: native (as per a thread in forum). Disk type is Hard Disk (scsi0) with VirtIO SCSI (Shall I try sata?) You can see logs in attached files. There are some following...

One node logs this error and vms are not reachable: VM 221 qmp command failed - VM 221 qmp command 'query-proxmox-support' failed - unable to connect to VM 221 qmp socket - timeout after 31 retries Could this be vm is eating so much resource or something related to pve? We have all updates...

When I use mail.domain.com and check "Use mx", it delivers fine. But how can pmg knows to use mail.domain.com even with higher priority? Also I can also uncheck "Use mx" when using mail.domain.com and it still delivers fine.