Website Rollback - Latest Info

[Steve Herschbach]

15 minutes ago, Rattlehead said:

What is the server load at when you experience these delays? I’m sure IPB has a way to display it either in the Admin CP or in the footer. Most forums do. Checking that yourself could at least confirm or rule out high server load.

As has been said it is being looked at.

[Hunterjunk]

Yep , still some strange little hangups every now and then .  Steve it's awesome the effort you have put into this for the benefit of all of us .  While life would go on without the forum , it would be the poorer for the loss . If it ever comes down to raising some funds don't be shy . You can put me down as a certainty , and I reckon most others here too .  

[CarsonChris]

Thanks for all the hard work Steve. I'm sure you'd rather be out beeping!

[Steve Herschbach]

Well day three, I’m starting to breath a little easier that the worst is behind us.  

[palzynski]

On 6/7/2024 at 9:46 PM, Steve Herschbach said:

 

...If the ninnies that charge me monthly for the software and the ninnies that charge me for the hardware can't tell me what is going on and fix it, then it's up to this little ninniest of ninnies to somehow make it work? It worked before and they are the pros, so fix it and make it run again. Is that too much to ask?

...

Normally a DB software has a perfo analysis tool which allow to diag perfos issues . These are a lot of reasons why lags can occur ; they can come from the DB itself ( indexes , etc ) , or from the HW ( disk , network ).

I dont remember the Oracle analysis perfo tool name but it allowed us to diag a major perfo issue . The reason was the DB RAID disk system incorrectly configured plus DB datas fragmented on disk . 

I remember an other perfo issue which was an Oracle bug on a specific SQL query . The dev team bypassed it quite easily by changing the query syntax .

I hope Invision has such perfo analysis tools , so that your support can understand and fix this issue ...

[phrunt]

SSD's are common place now, fragmentation is a thing of the past.   Invision gives plenty of performance information to base decisions on.  

I think all is good now, Steve's been on top of it all and doing all the right things.

[Steve Herschbach]

The server will be going offline for a few minutes for a memory upgrade. It will happen so fast most of you will not notice it.

Edit - Done!

[phrunt]

Working great for me, flawless  No problems at all.

[Steve Herschbach]

Well, I won't claim everything is fixed but the memory upgrade does seem to have helped. Still a few "stalled out and paused" type happenings but they seem to have lessened throughout the day. Still not perfect but if things stay like they are I'm happy and will not pursue anything more radical like a full move to another service.

Just an FYI backups occur each night at 10pm Reno time, so worst case in another meltdown at any time would be a rollback to the last 10pm backup. It seemed the crashes occurred in the middle of the night so I figure if everything is fine in the morning we are good, if not I can load the backup with little activity lost. Other than that I think we have made it through all this alive and ultimately for the better. Thanks everyone for your patience in dealing with all this, it is much appreciated.

I'll leave the announcement/notice up for a couple weeks for people who have been away and missed the fun, and then it will go away.

Steve Herschbach