Server Downtime

[lumpy]

I WOULD BE WILLING TO PAY AN ANNUAL FEE, I THINK MOST PEOPLE HERE WOULD BE WILLING TO PAY A REASONABLE FEE SUCH AS $10-25 PER YEAR. IT WOULD HELP WEED OUT ALL OF THE GUESTS.

[Ripple]

BC SHOULD GET A POLL GOING..BUT I THINK ITS GOING TO HAVE TO BE MORE THAN 25 A YEAR THATS WHY I SUGGESTED AROUND 100 ONE TIME FEE OR 25 IN 4 PAYMENTS OVER A FEW MONTHS ..

[BettorsChat]

Originally posted by frankb03
I would seriously consider firing whomever managers your server. These down times are inexcusable.

First, if a concern was preserving data then why wasn't daily backups done? Also, backups can be done with users on the server. If daily backups were performed the downtime today would have been decreased because an incremental backup could have been perfomed.

Second, one would think with a site of this magnitude RAID 5 would have been employed with hot swappable drives and redundant HW?

My servers are RAID 5 with four hot-swappable HDs. The cost is relatively minimal compared to being down for hours. In the long run it cost you much more. I do daily backups with a two week cycle. I also do daily backups of critical databases half way through the business day.

I don't know how to do daily backups, but will make sure that a cron job is setup to fix that. The place that i'm with if you want management issues you have to open a ticket as it doesn't come with the price of the storage and bandwidth. In other words these are not managed servers and if they were it would be way too high.

This site doesn't have Raid 5 its Raid 1. One of the drives completely went out Friday too so they had to put in a 2nd drive and move everything off my other drive onto it so I would be back at raid 1. Now i'm not sure, but the 2nd drive I have may be going out as well.

Reason they had to shut the site down Friday was they had to actually take the server and work on it on the bench because of the hardware failure. The SCSI drive wasn't the only thing that went out. Wherever the drives are normally placed that piece of hardware went out as well. Now the drives are on a backplane and it's going out as well. I was told if it fails I would lose all of my data on the 2 drives in there now. So they installed an IDE drive which meant taking the server down for 10-15 minutes. Tried doing a backup with it but kept getting errors. Eventually what happened was they had to shut down mySQL so they could back it up as that prevented any errors. Also if mySQL is running and you have people posting too mySQL I don't believe you can back it up while it's on. Now if you have a cron job set up late at night i'm sure that wouldn't be a problem.

Also the server went down as there was a Kernel panic Friday. Then yesterday it went down again as it locked up.

Basically i'm going to have too talk to them on Monday and see what all on this server is bad and going out for sure. Then see what my best remedies are, meaning to fix it or just buy a new server. I probably will replace the majority of the server except for the Ram, SCSI card and possibly the hard drives since they have 5 year warranties on them.

Another thing is that the NIC card is possibly going out. I will find out Monday if its still throwing out packet losses and collisions. They changed the duplex setting, but if its still doing the above then its no good. And they tried installing 2 NIC cards to bypass the one built into the motherboard with no luck. Now the original guy that told me that he thinks the NIC card is going out said usually that means theres a motherboard issue. So if that's the case I have no choice but to replace the server.

[ChuckLazar]

Hey Monte on other sites I see you can't look into discussions unless you register. You may want to think of that also.

[BettorsChat]

New Backplane fedex overnight was installed a little while ago. Now all I have to do is get Seagate to replace my drive that went out, but everything should be good to go.