I don't fully understand the cost structure for server hosting, but it may be worth having a discussion about moving to another host and what that would cost.
You don't realize how many times that discussion has already happened. It's easy to take hosting services at their word about what they offer in their ads and information, but the problem is what actually happens once you've paid them for the year. THAT is what this is, the FO afterward. This host is on 2 continents (UK and US) with backup redundancy, AND is using AWS infrastructure as of about a year ago... yet this is what we go through every few weeks or a month. And because it happens that VHT is hosted on their server in the UK, I can't just call them to ask WTH is going on. Here is a transcript of my fun with them for the last month.
[Note: the interface with the host, copies of which you see below, only allows plain text. In order to show how pissed I am,
I can only use capitals - any italics or bold text you see in my reverse copy/paste below are done here for emphasis]
------------------
#304577 - Site down
Tom H.
June 26, 2026 at 6:05 PM EDT
As I'm sure you are aware. What was it this time? Any estimate of when things will come back up?
Billy
June 26, 2026 at 6:34 PM EDT
Hello, The server was rebooted, but it's not down. I'm investigating the issue a bit more with you on this and will update you shortly. Thanks! ---- Billy
Tom H.
June 26, 2026 at 7:14 PM EDT
It's back up, hopefully for good because it's running better than it has in quite a while. Thanks, Tom
Ben
June 26, 2026 at 7:19 PM EDT
Hello Tom,
We found that the resources allocated to the account reset after we applied our urgent security fixes today.
[AD - remember this]
Apologies for the delay but it should be stable from here. Please let us know if you encounter any additional issues. Thank you! ---- Ben
Tom H.
June 27, 2026 at 7:20 PM EDT
Referencing the mention of resources allocated to our account in the previous reply by Ben,
would it be possible to have notes about that in our file so the next time there's an outage or reset or whatever happens, we can get it solved more quickly? That said, we have a different (but possibly related) problem in the last half hour. I'm getting mass quantities of server error messages that return as fast as I can clear them, all the same about too many connections. Prior to our outage yesterday I was able to run the timeout length at between 40 and 60 minutes regularly with few if any errors. Now this afternoon, even with the timeout set at only 5 minutes I'm still getting hundreds of server errors in just a few minutes. As of this very moment there are over 1000 guests and only 5 members online with timeout set to 5 minutes. We never had that problem prior to yesterday with an average of 1500 guests any given day along with 15 to 30 members, so something has changed. Observations? Tom
Tom H.
June 27, 2026 at 8:06 PM EDT
And it just came to my attention that our mobile version shows "Oops! We ran into some problems" on any page of the forum. Logged in just fine, can't go anywhere including Home or private messages.
Billy
June 27, 2026 at 9:01 PM EDT
Hello,
I would restore the files back to before this oops message showed up, or see if you can locate someone who specializes in XF. [SERIOUSLY?]
I'm not seeing a server level error for this particular issue. The other issues were related to memory level, I went ahead and adjusted those again, but it's not helping with the oops! problem. Thanks! ---- Billy
[AD - at this point I'm wondering if anyone but Jeremy actually knows what they're doing there, and I'm NOT doing what he suggested]
Tom
June 28, 2026 at 8:04 PM EDT
Right after I sent the last message about the "Oops" page in our mobile last night, it suddenly worked fine. I took a shower after the forum was working pretty well today (
aside from 10s of 1000s of server errors despite me having the timeout set as low as 5 minutes for quite a while) and come back to find this attached screenshot taken 3 minutes ago. I can log in, but just like the mobile was last night, I can't navigate to anywhere. Can we get this fixed?
(sent screenshot of Oops! page in every section of the forum)
Billy
June 28, 2026 at 8:41 PM EDT
Hello, I'm going to have to escalate this over to our engineering team so they can take a deeper look into the issue. we'll update you once that's been reviewed. Thanks! ---- Billy
Ben
June 28, 2026 at 9:39 PM EDT
Hello Tom, Whatever errors are being logged are not being logged in a way that's visible to us. We're happy to take a closer look, but we'll need administrator access to the management area. Can you please provide us with the appropriate credentials? ---- Ben [yeah, letting someone completely unknown into the backend of VHT feels so safe. Not in need of any more ridiculous suggestions, I waited until the next day which was Monday]
Tom
June 29, 2026 at 10:29 AM EDT
Good morning (here), I hope Jeremy is working today, he knows what we've been through over the last 6 months to a year and the tweaks that made things very manageable prior to last week's outage that has changed a LOT in the behavior of my forum. I'm having a problem with the fact that prior to the outage, we could have our timeout set at 60 minutes and despite that allowing anywhere from 1500 to 3000 guests at a time while 20 or so members were logged in, over the last few days I've cut that timeout down to 5 minutes and STILL had to battle 3000 to 6000 bots while only 2 or 3 members were shown logged in. I need it explained to me why the attached screenshots are what is happening now, when prior to the outage things were very different and far more manageable. And it was all solved without the need to get into my admin area. AND we didn't need to restore any files back to any specific time ever.
With only 4 members and only 57 guests logged in, please explain the massive amount of server errors generated in less than one minute when the screenshot was taken (after I cleared the errors for the millionth time over the weekend)
that previous to the outage were never generated by the same exact (or far greater) conditions.
Sunil
June 29, 2026 at 10:45 AM EDT
Hello, Thanks for your update. Jeremy will available later today and for further updates please standby, thank you. ---- Sunil
Ben
June 29, 2026 at 12:21 PM EDT
Hello Tom, Unfortunately Jeremy is away this week. I'm looking into your issue and will provide an update once I have more information. ---- Ben
Ben
June 29, 2026 at 1:49 PM EDT
Hello Tom, Based on the errors you provided I verified the xf_session_activity table was filling up. This is a known issue with XenForo when you see a large volume of traffic in a short period (bot or legitimate) and it fills the table to the database server's maximum "MEMORY" table size until the cleanup script runs. This is also why the user count appears inflated, because hosts are establishing a connection without logging in and artificially inflating the count. I have substantially increased the "MEMORY" table size limit, but please monitor and let us know if you encounter any new cases. ---- Ben
Tom
July 13, 2026 at 10:26 AM EDT
Down again. The outage a couple weeks ago on a Friday resulted in a lot of problems the following week, most of which were eventually cleared up. But yesterday things started getting really slow to respond and this morning it continued until it's now down completely.
Tom
July 13, 2026 at 10:35 AM EDT
And it's back up, but still sluggish like yesterday. It would really be nice if we could get more than a couple of weeks worth of nice, steady, consistent uptime without sluggishness and outages.
Jeremy
July 13, 2026 at 4:31 PM EDT
Hi Tom, Thank you for the updates, I'm sorry to hear you have had some recent issues on the site that have resurfaced. I've navigated around your forums but it is lightning fast for me here. Does the slowness occur during specific site actions and if so can you provide us any necessary details to replicate the issue (including any login details required)? ---- Jeremy [HOW MANY TIMES MUST I TELL THEM WHAT IT'S DOING WHILE SENDING MULTIPLE SCREENSHOTS??]
Tom
July 16, 2026 at 10:52 AM EDT
Jeremy, the slowdowns happen in many forum sections, but mostly in the What's New link and sometimes in the messaging system too. When I go to a new post in the What's New link, which I visit hundreds of times a day, it sits there with the little icon spinning in the Chrome tab and often takes as long as 10 seconds or more to actually go to the new post, same with my PM link. Click and wait. But it varies throughout the day, sometimes it reacts almost immediately only a few minutes later so it is not consistently slow. This is what is SO annoying when from your end all seems fast as a rocket, while I'm slogging through 'mud'. Screenshots don't show much but I attached one from day before yesterday, right during it happening on the What's New link when I clicked the Likes menu.
Jeremy
July 16, 2026 at 5:42 PM EDT
Hi Tom, Thanks so much for the detailed feedback and for your patience. I have made a tweak on our side which I hope remedies this; can you do some further testing and let me know if the behavior persists? If it does, would it be possible to provide us a login to your forum so we can try and replicate the issue and do some further testing? Also, the next time this happens can you try and record the approximate time and the exact URL you were visiting? ---- Jeremy
Tom
July 16, 2026 at 6:24 PM EDT
That screenshot was Monday at 7:25 PM EDT, and the link to the page was in the address bar which I figured you could see.
vintagehondatwins.com/forums/whats-new/ is the page I was on vintagehondatwins.com/forums/posts/206019/ is the link to the "Like" I was trying to check, it's shown in the lower right of Chrome (hover url) in the screenshot. After clicking on that like, it literally took over 10 seconds to respond. It also happens fairly often on my PM Inbox as well, and that address is vintagehondatwins.com/forums/conversations/. And it also happens a fair amount of times I try to refresh the Who's Online page, that address is vintagehondatwins.com/forums/online/?type=member. I will say that it does seem improved after whatever it was you did, but off and on during the day it improves and then it deteriorates again. Ever since that roughly 2 hour outage a couple of weeks ago right before you were gone for a week (Jeremy), things have been quite chaotic, beginning with massive issues with guest/bot totals immediately after the site came back up so I reduced the timeout down a bit at a time until it was down to 5 minutes. But we still had multiple server errors regardless how short the timeout was set, and we literally hadn't had those problems for many months leading up to that outage (I don't care if it was characterized as a reboot and not an outage, our site was down for close to 2 hours).
Jeremy, this below is from June 27th, perhaps you missed it. -- Tom
Referencing the mention of resources allocated to our account in the previous reply by Ben, would it be possible to have notes about that in our file so the next time there's an outage or reset or whatever happens, we can get it solved more quickly? [Please? Because I'd hate to have to go through that again while waiting for all the tweaks to be reinstated.] THIS below: That said, we have a different (but possibly related) problem in the last half hour. I'm getting mass quantities of server error messages that return as fast as I can clear them, all the same about too many connections. Prior to our outage yesterday I was able to run the timeout length at between 40 and 60 minutes regularly with few if any errors. Now this afternoon, even with the timeout set at only 5 minutes I'm still getting hundreds of server errors in just a few minutes. As of this very moment there are over 1000 guests and only 5 members online with timeout set to 5 minutes. We never had that problem prior to yesterday with an average of 1500 guests any given day along with 15 to 30 members, so something has changed. Observations?
Jeremy
July 17, 2026 at 1:01 PM EDT
Hi Tom, Thanks so much for the additional details. We didn't have any server side issues on Monday and this may be a case where additional optimization is required at the application level. I understand that you are the developer for this particular site/forum, so you may need to do some reading up on additional tweaks you can make to XenForo to help speed things up further. One option you can try is enabling debug mode in XenForo the next time you face a similar performance slow down and this will give you a breakdown of execution time, queries, etc. You can also check any addons you have enabled in XenForo and disable those one at a time to rule out any culprits. If you have not already done so, then enabling Redis for object caching in your XenForo configuration would also help with these types of slowdowns. ---- Jeremy
Tom
July 17, 2026 at 4:57 PM EDT
"I understand that you are the developer for this particular site/forum"
I guess you could say that, but I'm more like the guy who knows computers fairly well and happen to be the one who built VHT with some help. I'm only barely equipped for what I do, but I manage. The term developer is far beyond what I'm capable of.
"One option you can try is enabling debug mode in XenForo the next time you face a similar performance slow down and this will give you a breakdown of execution time, queries, etc."
I'll need guidance for doing that since I don't want to break anything and it's a sensitive area. I have a couple of members who should be able to help. And it would be better to have some idea of what causes it.
"You can also check any addons you have enabled in XenForo and disable those one at a time to rule out any culprits."
I'd be surprised if one of my 4 total add-ons might be partly responsible, all of them have been in use for years and only one came from someone other than Andy on the XF forum. And all of them work well, no glitches or error messages.
But again, can't the tweaks currently adjusted on your end for VHT be documented, so the next time there's an event that causes some reset leading to bizarre behavior that clearly was not caused on my end, it can more easily be readjusted without repeated requests? That would help immensely. Thanks, Tom
Jeremy
July 20, 2026 at 3:31 PM EDT
Hi Tom, Thanks so much for the additional information and clarification. If you would like to try enabling debug mode to get further insight into a performance issue if/when this resurfaces, you can open the "src/config.php" file for XenForo using the cPanel File Manager and add this line to the bottom: $config['debug'] = true;
Alternatively we can add this for you on request, just let us know. To disable debug mode you would only need to remove that one line from the above file. I have also added a note to this ticket (#304577) with the tweak made for your account, so if you have any problems you can reference this ticket ID and any support member on shift will be able to quickly review and make sure that is in place for you. ---- Jeremy
Tom
July 24, 2026 at 9:25 AM EDT
Well, here we go again. Yesterday afternoon we had yet another brief outage, and once again afterward things are sluggish. I waited to see if it would improve on its own but of course, it didn't. Tracking what caused it would be nice, sure, but I just want to know why it ALWAYS happens right after things get so sluggish that it generates a couple of minutes of "server down" messages, and why it's always just like it is right now once it comes back up. SLUGGISH everywhere. While I could try those things you've suggested, what bothers me most is in between (after I complain and the tweaks are made yet again), things run fantastically - every single page displays in less than a second, page changes are nearly instantaneous, all functions work as they should regardless which section of the forum I'm using, all while ABSOLUTELY NOTHING HAS BEEN CHANGED ON MY END. Nothing - and it can go on for days working perfectly after I've done nothing on my end. UNTIL it happens AGAIN due to something NOT ON MY END. This is the point I've been trying to make for months. I change nothing, yet once SOMETHING has been done on your end, then everything is great until it happens again. And when it DOES happen again it is NOT caused by something we've done -
because nothing was changed. I'll go along with my day as normal, praising the good work done on your end in my head and thinking "this time it's really fixed" and then I get the same f-ing disappointment all over again, along with another outage usually, and it's rinse and repeat. And I'm REALLY tired of it.
Jeremy
July 24, 2026 at 10:50 AM EDT
Hi Tom, Do you happen to have the approximate time that you experienced that brief outage? I ask because we did receive an important Linux kernel security update that we applied at Thursday July 23 00:56 UTC. There would have been a few minutes of downtime associated with this while the server was rebooted into the new kernel. Normally we use KernelCare to provide rebootless security fixes for the kernel, but KernelCare patches were not available in this case and since it was a critical LPE vulnerability we needed to act fast to ensure optimal security for all of our users. ---- Jeremy
Tom
July 24, 2026 at 11:18 AM EDT
In my frustration of the moment, of course I did not notice the exact time. But it does sound like it coincided with the reboot. But my issue is, after EVERY incident like this we go through the same thing again. From working like normal, quick page loads and normal transition speeds, to sluggish behavior where it takes a page sometimes 30 seconds or more to load, or for a PM to be sent, or for a post to be applied to a thread, you name it, it affects the speed of everything IMMEDIATELY AFTER these little random outages caused by needs on your end. AGAIN, what bothers me is that it's so consistent - after every little outage it's as if settings are lost and we go through this repeated attempt to identify what and why, fruitlessly, because on my end WE'VE CHANGED NOTHING TO CAUSE THE EFFECT. I'm quite sure you would also be thoroughly annoyed if the same thing were happening to your site every week or so. The worst part about it from my perspective is, we're paid in full to next year and we're stuck with it regardless. And that is not a position I care to endure without some sort of progress. You're always explanatory and diplomatic Jeremy, and it's greatly appreciated, but that only goes just so far when the behavior is repeated ad infinitum.
Jeremy
July 24, 2026 at 3:36 PM EDT
Hi Tom, I totally get where you are coming from and unfortunately it has been a very hectic month on the security front with many Linux local privilege escalation vulnerabilities being identified by AI and requiring way more reboots than we typically would like. However, security is paramount to us and being on a shared hosting environment I'm sure you can appreciate the work we have been doing to make sure that your server remains secure. We're hopeful that the recent stream of Linux issues (which is extremely abnormal in the history of Linux) will wane soon and we're also putting pressure on CloudLinux to release KernelCare patches more quickly so we wouldn't have to reboot the servers to provide these security fixes. I can tell you that we are doing everything in our power to minimize these disruptions and we do appreciate your patience. ---- Jeremy
---------------------------
I do NOT feel comfortable with someone I've never spoken to having backend access, but that's just me. ESPECIALLY when no matter what we do here in terms of documentation or minor adjustments, the software for this place hasn't been touched, there are zero file errors every day during automatic checks, and I'm constantly monitoring the server timeout based on guest totals to try to keep things happy. We change nothing but every single time we have any length outage, it's rinse and repeat.
