New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
moved to a german shared hosting service (that i use since 2002)
way to much. and it felt like prices were being raised continuously.
I am using Apple Mail on MacOS, not mobile (well mobile too, but not too worried about that one) and below is straight from the horse's mouth:
Let me know if it happens anymore from here on and on which days. I'll see if it lines up with any upchucks/restarts we do. I got some other ideas related to 'long living TCP connections' which might be set to timeout too fast.
Francisco
My employer requires me to use Gmail. My experience with workplace Gmail doesn't even come close to Fastmail. So if zohomail and inbox.eu are being mentioned in the same breath then I'm naturally skeptical, but maybe they deserve a fair shot.
I'm curious to know what complaints can be made against Fastmail (other than pricing).
indeed us1
it seems better now, but honestly it's ok, no biggy, thanks for caring.
it seems better now, but honestly it's ok, no biggy, thanks for caring.
For sure been better with the changes we made last night
Thanks for the confirmation.
Francisco
N/A
I'm still seeing outages of the US IMAP host, no longer the usual nightly 5 - 15 minutes outage, but they now started being more "distributed" at different times of night and day. For example today there were 2 reported - one at 1:45PM and another at 7:30PM EST. Yesterday another one at 2:00PM. There was also one on Mar. 18 at 10:10PM EST
And there was another 5 minutes glitch last night (3:40 AM to 3:45 AM). Not complaining, I haven't seen it since about March 10 (aside from what I reported on Mar. 26)
And another 12:10 PM to 12:15 PM today. Might be just a glitch, and it's definitely not a regular nightly occurrence like before
I didn't see either of these on hetrix, let me know how you're monitoring it.
We did have one this morning but that was just an OOM. Over time smartermail just gets very fat (200G+ on main process) so either we punt it, or it just gets OOM killed on its own and restarts. It might be a decent idea to just do a scheduled reboot once a week.
The nightly throwups were fixed by SmarterMail a few versions ago so thats been nice
Francisco
Service monitor on us1.workspace.org [port 993]:
https://hetrixtools.com/report/uptime/5dc77cbdf7b7e13285a987a64bcf8e73/
The actual outage could be much shorter, but I'm only sampling every 5 min. Let me know if that's a burden on the server.
https://hetrixtools.com/report/uptime/5dc77cbdf7b7e13285a987a64bcf8e73/
I'd have to add one for 993. Our probes are on port 25 & web side and those have been the most reliable signal if there's something acting up. Its possible SM restarted the IMAP part of itself, I'd have to check a log.
Thanks for the link
Francisco
Could you possibly comment on the other thread where I asked about the grey-listing?
Thanks!
If you're getting OOM, your software has a bug, your configuration is wrong, or your resources is insufficient. It does seem like your company runs into that a lot and just reboots/restarts shit and that's why there's so many frequent downtimes logged.
Its just how smartermail does their caching. Over time it caches a lot and it'll OOM. We've thrown hardware at it which has helped a lot, and maybe some more tweaking in how RAM allocates will help more so. The Vegas boxes have 768G RAM, with half ish of that going to ZFS.
There's threads on their forums about it, and they've merged a lot of fixes that have helped. If you wanted you could check their changelog and they mention small memleak fixes a decent bit the past few months.
I think a predictable once a week reboot off hours would be perfectly fine, since they put out software updates on the regular (every couple weeks, sometimes once a week) so we'd just group it with that. We try to stay at most 1 version behind (if there's a lot of changes we'll hold off to let it cook for a bit).
Francisco
No, that means the software is crap. You don't work with disciplined enough developers. I can already hear the rants from two developers I've worked with who would just rip your attitude and smartermail apart.
If you regularly OOM, you're doing it wrong. Full. Stop.
Hi @Francisco, can I ask you to please take a look (or have someone do it) at Ticket 22255? I've asked a few pointed questions, and instead of answering those I got an unsolicited suggestion ("use SpamExperts") and now grey-listing has been turned off on my domain without my asking for it. I want grey-listing active (I realize you're not a fan) and use the "trusted senders" option to experiment with in order to reduce delivery delays.
You also said you're looking into exposing the grey-listing control in the admin email user panel - any chance of that happening soon(ish)?
Thanks in advance,
PS: there was another short outage today @ 4:55PM EST
Edit: and another just detected @ 11:50 PM
All of my Cranemail-hosted domains are down as well.
I don't understand pricing for inbox.eu. What is account and how many domains it can host on single account?
Sorry about that.
I did a couple reboots to fix an issue where domains weren’t attaching properly when … reboots would happen. Was a recent issue, I think likely related to some extra validation SM does.
It seems to work clean now though so we will see in a week.
I was hoping doing the restart in the middle of the night would be ok, but I guess the 1 or 2 minutes was still seen
If you’re still having issues PM me the domains to look.
Francisco
1 user = 1 inbox, if that's what you're asking.
no clue how many domains
Yeah it was pretty rough on mobile back then worth another shot now though, they’ve pushed a bunch of fixes since
As far as I know, you can add an unlimited number of custom domains. You pay per user.
Not sure what inbox.eu has to do with this topic...
someone did recommendation, not just inbox.eu but also other services.
So I got a very succinct reply to the ticket:
1. Our Smartermail controls the graylist
2. The default setting is 15 minutes
This raises another question - why would the recipient server (NameCrane SmarterMail in this case) have any say over the retry timeout of the sender? Wouldn't the sender be the one to determine when to retry?
PS: @Francisco, I'd love if you could expose some more SmarterMail controls in the domain admin / user control panel (greylisting included)
@Francisco
Is there a feature request form/email for CraneMail?
I would love to see something like Addy.io (or even SimpleLogin or Proxiedmail) to create and manage aliases (and track or block if needed).
PS: Interesting to see how name contractions work from language to language. People here usually go for the gender neutral FRAN for Francisco, my mind goes to Chico as I grew up with that as most common contraction....
>
I don't even think Fran is a proper contraction of Francisco, it just.... happened basically. Keeko, Pancho, Chico, are supposed to be the the usuals.
That's been requested many times, its high on the list and will work on it after domains/newbilling launch
Francisco
fernando