What happened was that some email was piped prematurely to the main server since qmail was turned on for 5 minutes to get the data restoration to start.
Some email was bounced back since all directories were not yet created.
Error on our part, I am sorry. Not all, but a few was bounced.. Not happening anymore
If you have valid questions, please post them and maintain professionalism while posting here.
We have an experienced admin team dealing with issues as fast and professionally as possible. Please be patient while we recover.
Yelling or shouting isn’t going to make things faster.. If you are not satisifed with our services, you are free to look elsewhere. We at Jodo do our best to maintain high standards of service
Yash, could you please give me an estimate of how many of the stored emails on the backup server got lost/bounced back during those 5 minutes of uptime of the mainserver wehin it still hadnt the accounts configured?
All of them? Something like 10%? Half of all? After all, there is no use in having such a backup server storing mails if they get lost anyway, some hours later. This way, nothing got better since last major email issue, regarding backups.
Thank you for the quick answer.
10 to 20% of the emails of each domain, or 10 to 20% of domains lost all their mails?
Yes, correct, the sender gets an error message and has to resend. But they are lost in terms of me knowing they ever existed, if they do not resend, thats what I meant.
12GB extracted. 5GB remaining. That should take 40 more min
After that, phase 2 begins which means updating user directories with extracted data. ~ 1 to 1.5 hour
Sorry for the upward revision. Doing everything we can speed this up
I understand this incident has hurt our reputation. We did quite a bit of preparation for mail server disaster recovery and ensuring 100% reliability. But having a RAID failure is the WORST thing that can happen to a hosting company
I am already discussing having a hot backup instead of a standard backup server with our team… See how we can implement this.. as soon as we get this over with
Please give Jodohost sometime to do their job properly and recover the mail server so that imcoming email are not lost as these things do happen from time to time. I’d appreciate if Jodohost can ensure their customers that they have done everything they possibly can to prevent such an event do not occur again with any of the servers.
I dont know your situation, but for me and all of my customers, the email is a tool work. So, if we doesnt have that tool then we lost money. That simple. I understand the extraordinary thing of this issue (and i know, and its posted, their efforts to solve it). Meanwhile, im losing money.
You will receive the spam… the email hasn’t gone anywhere.. Just the primary mail server is still down
At the moment, data is being pushed back into their directory. 7GB out of 15GB done. I’d say about 40 minutes for the remaining GB
Now the problem is that 2GB of data wasn’t extracted out of the backup file. we have two choices to do extract that 2GB…
We pass a command asking the tar to ignore files already extracted
We do a compare against the files already extracted and ask the TAR to exclude those
Extracting the remaining 2GB would be the quickest.. but ensuring that it gets to those 2GB the quickest is the challenge. Our guys are on the test servers figuring it out
I am extremely sorry for all this. The recovery has taken longer than expected, IO is bogged down quite a bit.. even on this high-end server. I can’t imagine thinking how long it would have taken on a SATA server.
As soon as we are done with this, I am going to be putting up a new mail server and have our guys work out a way so we can have an active data sync so that if a mail server ever goes down again, we can simply swap roles..
Little stressed out here, dealing with alot of angry customers. I am going to take a 15min break while the admins handle it..
This is an automatically generated Delivery Status Notification
Delivery to the following recipient failed permanently:
[email address removed]
Technical details of failure:
PERM_FAILURE: SMTP Error (state 10): 553 sorry, that domain isn’t allowed to be relayed thru this MTA (#5.7.1)
I can’t post a ticket because I can’t get to my control panel (site unavailable). I’ve tried to contact Live Support but was forwarded to post an email instead… Sending to [email protected] seems to be a black hole. I’ve never gotten a response or confirmation when using this address (and I’ve had to as both the reseller CP and my CP have been down frequently as of late).
I’m patient with alsmost all issues, but when emails are bouncing, I start to get very anxious.