Background... Server raid failed, rebuilt raid ran chkdsk but now I am
unable to run SQL service.
I've tried to manually start service but receive the following message
The MSSQLSERVER service on the local computer started and then stopped. Some
services stop automatically if they have no work to do.
Here's all I have in the error logs
2007-08-19 10:56:39.98 server Microsoft SQL Server 2000 - 8.00.2039
(Intel X86)
May 3 2005 23:18:38
Copyright (c) 1988-2003 Microsoft Corporation
Developer Edition on Windows NT 5.2 (Build 3790: Service Pack 1)
2007-08-19 10:56:39.99 server Copyright (C) 1988-2002 Microsoft
Corporation.
2007-08-19 10:56:39.99 server All rights reserved.
2007-08-19 10:56:39.99 server Server Process ID is 1812.
2007-08-19 10:56:39.99 server Logging SQL Server messages in file
'f:MSSQLlogERRORLOG'.
2007-08-19 10:56:40.01 server SQL Server is starting at priority class
'normal'(2 CPUs detected).
2007-08-19 10:56:40.15 server SQL Server configured for thread mode
processing.
2007-08-19 10:56:40.18 server Using dynamic lock allocation. [2500] Lock
Blocks, [5000] Lock Owner Blocks.
2007-08-19 10:56:40.23 server Attempting to initialize Distributed
Transaction Coordinator.
2007-08-19 10:56:42.29 spid4 Starting up database 'master'.
2007-08-19 10:56:42.31 spid4 Error: 5172, Severity: 16, State: 15.
2007-08-19 10:56:42.31 spid4 Error: 5173, Severity: 16, State: 1.
2007-08-19 10:56:42.31 spid4 Error: 5180, Severity: 22, State: 1.
I have a server setup with the standard recommended RAIS(10-5-10 setup (10 for the OS, 5 for the data, and 10 for the trans logs). Running out of space on my RAID 5. Have lots of extra space on my RAID 10 where my trans logs are. I currently dump my files to disk and then use tape to back them up. I have been putting these files on my RAID 5 array, but was going to move them to my RAID 10 array. Anyone seen any downside to doing this?
Quick question in setting up a 3-disk SQL 7.0 system - can anyone think of a benefit to segregating a single RAID 5 disk array into numerous logical partitions for separating out the OS, the database files and the transaction logs? I would assume performance would be unaffected (as the drives are acting as a single array for reads & writes anyway) so other than general organization what (if any) advantage would be gained over making a single large logical partition?
Have a problem with SQL today. I rebooted the server and now it cannot see the E: drive, as per the RAID array setup. Problem being is that all my SQL databases and logs are on this drive. I'm guessing that the error message is just misleading as based on Microsoft this means the server is stand alone and needs to be started with a non-Domain account. Fix the hardware RAID array and hope all is well. Does that sound about right to you?
I am configuring a new database server, without SAN access, and want to know what is the best practice for SCSI RAID configuration. Do most folks prefer RAID 5 or RAID 10 configurations where their databases will reside?
RAID 5 beats RAID 10Can I get some feedback on these results? We were having some seriousIO issues according to PerfMon so I really pushed for RAID 10. Theresults are not what I expected.I have 2 identical servers.Hardware:PowerEdge 28502 dual core dual core Xeon 2800 MHz4GB RAMController Cards: Perc4/DC (2 arrays), Perc4e/Di (1 array)PowerVault 220SEach Array consisted of 6-300 GB drives.Server 1 = Raid 103, 6-disk arraysServer 2 = Raid 5 (~838 GB each)3, 6-disk arrays (~1360 GB each)TestWinner% FasterSQL Server - UpdateRAID 513Heavy ETLRAID 516SQLIO - Rand WriteRAID 1040SQLIO - Rand ReadRAID 1030SQLIO - Seq WriteRAID 515SQLIO - Seq ReadRAID 5MixedDisktt - Seq WriteRAID 518Disktt - Seq ReadRAID 52000Disktt - Rand ReadRAID 562Pass Mark - mixedRAID 10VariesPass Mark -Simulate SQL ServerRAID 51%I have much more detail than this if anyone is interested.
We have (had) an active/active cluster. 2 physical machines, each running their own instance, clustered together. Node1/Ins1 and Node2/Ins2.
Node2 failed and Ins2 failed over to Node1 as it should. Node2 required that we rebuild the server (rebuild = reinstall O/S). Now we need to get Node2 back into the cluster and get Ins2 failed back over to Node2.
Does anyone know, for certain, the correct steps to accomplish this? Obviously, we could backup everything and completely destroy Ins2 and recreate it on Node2 then rejoin the cluster. But I'm looking for something less destructive.
Is it possible to reinstall SQL, then rejoin the cluster, and then fail Node2 over? Or will there be registry conflictions?
Any help would be appreciated. Also, if you have any links to some official documentation, that would be great too.
I have a stored procedure that has a paramter that accepts a string of values. At the user interface, I use a StringBuilder to concatenate the values (2,4,34,35,etc.) I would send these value to the stored procedure. The problem is that the stored procedure doesn't allow it to be query with the parameter because the Fieldname, "Officer_UID" is an integer data type, which can't be query against parameter string type. What would I need to do to convert it to an Integer array? @OfficerIDs as varchar(200) Select Officer_UID From Officers Where Officer_UID in (@OfficerIDs) Thanks
I have a requirement to only rebuild the Clustered Indexes in the table ignoring the non clustered indexes as those are taken care of by the Clustered indexes.
In order to do that, I have taken the records based on the fragmentation %.
But unable to come up with a logic to only consider rebuilding the clustered indexes in the table.
My SQL 7 is on RAID 5. Sometimes on non-peak hours, on RAID disks first two lights ( from left ) are constantly on for hours. NT Task manager, nothing unusual, SQL current activity - no running user processes. Isn't second light on RAID comes on if any disk activity ( Read/Write ).
I`ve tried implementing NT Software Raid / Stripping with Parity and am unable to stripe disc that are more than 2g and use SQL. I have not found any info in technet. Any ideas! Thanks.
Hi,I was going to buy a server with Raid 1 as I thought that it meant that ifone of the two mirrored drives fail, you simply take it out and put a newone in. At which point presumably the hardware takes over and copies theother drive over to mirror it again.However, my sql server admin book, says raid 1 is bad, as it means you havelots of downtime, when recovering from a broken drive.Can anyone give me some advice on this? What is the best Raid to use whenyou are running SQL server on the server.ThanksJJ
Im setting up a hardware raid 5 solution for one of our db servers. The data files will reside on the stripe. We dont realy want to raid more drives for the Transaction log if its not nessesary. If the drive with the log crashes is the data file for the database useless ?
Hello, I run a small homw office. I am planning to purchase a dell powerdge 1750 server to install SQL server on that. I am confused here about which RAID should I install on this server RAID 1 or RAID 5. The dell customer rep could not tell me the advantages of installing only RAID 1 or only RAID 5 or installing both RAID 1 and RAID 5
I am recommending that we change our Raid Configuration on some of ourServers from Raid 5 to Raid 0+1; we are experiencing severe IObottlenecks.Our hardware guys are pushing back a bit. They claim that Dell has aweird implementation of 0+1 and told me something about one drivefilling up before it begins to write to the next. They claimed thatthis gets rid of most of the benefits of 0+1.I know that 0+1 is not as good as 10 for availability, fault tolerance,and rebuilding, but shouldn't the write throughput be about the same?Setup:Poweredge 2850Powervault 220SPerc 4/DC Controller 1Perc 4e/DI Controller 0
I have always used RAID 5 for my database apps. I am wondering though is this the best solution. I am purchasing a new server and deciding whether to go with RAID 1 or RAID 5. This server will support all our .NET Apps, SQL Server 2005, Reporting Services, and Integration Services.
Server: Msg 3132, Level 16, State 1, Line 1 The RAID set for database 'xxxxx' is missing member number 2
This happens when I try to run a restore?? is it because my server is not set up as RAID 5, I had took a dump from one server and try to apply it to another? Suggestion please.
I heard that is said that MSQ SQL 2000 doesn't work correctly with raid (for example raid 5) I have my application in 2 companies. The first one with not a raid controller (even though the computer has a raid controller) executes a backup when no users are online in more than 8 minuter and the file of the database backup is 3,5 GB. The memory of the computer is 2GB. In the other company when more than 20 users work in the database with a raid controller the backup is occured in 2,6 minutes and the database backup file is more than 6 GB. The memory of this computer is 4 GB. What is your opinion? Is this rumor true? Do you suggest I should apply raid 5 to the first server? Regards, Manolis
We have a huge table with around 250 million records and have implemented SQL server 2005's new table partitioning feature. Now the data seems to be evenly spread across 20 different filegroups ( each 5 GB approx ) for the same table that was occupying 100 GB itself in the PRIMARY filegroup earlier.
Still the query response times have not come down drastically but we could see a good improvement in the execution plans now.
WE ARE USING RAID 5 IN OUR PRODUCTION ENVIRONMENT. ANY IDEA / THOUGHT ON HOW TO PLACE THE PARTITIONED FILEGROUPS AND THE LOG FILES IN THE RAID 5 (BTW , I'm very new to RAID concepts , any detailed instruction would be helpful ).
My company is in the process of upgrading our sql 2000 server to 2005 on new server hardware. In this process I'm charged with the actual server and database configuration on the new server so currently I'm investigating how to take advantage of new 2005 features and improve the use of common features on both platforms. We've decided to perform this upgrade in three stages (won't bother you with why) and the first stage is to install the new server with a 2000 installation with a different configuration regarding files/filegroups on the new hardware configuration that has a different raid configuration. The old server configuration was lacking in just about every department so I'm serious about getting it right this time. Its an OLTP system btw.
The question(s):
The 2780A course book states that sql 2005 can perform the following scans: "Multiple parallell scans for a single table if the filegroup of the table contains multiple files." Does this imply that the files must be on separate disks/raid channels? Ie, if I put all files in the filegroup on the same RAID 10 channel will I gain nothing over having just one file?
Can 2000 take advantage somehow of several files in a filegroup to achieve greater parallellism or is there no (end user) performance oriented reason to have more files in a filegroup? (Disregarding any impact of different backup strategies for several files).
I'm intending to place all nonclustered indexes in a separate filegroup with all files assigned to a dedicated RAID 0 channel for striping. Due to the lack of fault tolerance here I'm wondering if this is wise. But since nonclustered indexes are basically secondary metadatastructures and doesn't contain any data I'm wondering wether the server can disregard using an index when executing a query if the index resides on disk which have failed and the index is inaccessible. Is there a fallback option available to the server to perform a table scan or use another index instead?
Basically the same question can be asked for tempdb, use a RAID 10 with four disks or dare I use a RAID 0 with four disks (since 2005 use tempdb more heavily than 2000).
Also considering to put the OS on a dedicated RAID 1 and a dedicated RAID 0 for the pagefile, same concerns apply.
I appreciate any input, also examples (or links) or your experiences of good RAIDing practices for OLTP systems.
Thanks for your time.
PS. Sorry for posting this in both 2000 and 2005 forums but I don't wanna miss out on any good answers. ^^
I have a RAID 10 container consisting of 4 146GB, SAS, 3.5-inch 15K RPM Hard Drives.
We have the OS on its own RAID so we are strictly dealing with setting up SQL Server 2005 on the RAID 10.
SO ... is it better to create seperate volumes for the Data, TLog and TempDB file placements (i,e, C:, F: and G or is it alright to create just plain-jane folders for the file placements?
Ok - dont' throw rotten food at me for asking this question...
Is there any advantage (faster I/O) to creating multiple mdb files on the same RAID 5 for the same database?
In other words, database ABC is 100 gb in size and has one primary file on H:. If I create another (or 3 or 6) secondary files for the mdb, would there be an appreciable performance gain? Same with log file?
How to choose the right RAID level for a database server based on disk performance counters? Lets say I have the collected performance counters data for Disk Transfers/sec, Disk Writes/sec and Disk Reads/sec.
Or is there any other best method to decide the RAID level?
I have to spec out a new server, and I have the option of using RAID-5 or RAID-1 drive sets. I am limited to 24 drives in groups of 6, and I have to have one hotspare per group, so up to 5 usable drives per channel.
I need 80-100GB of space total.
Okay, you're waiting for the question....I have heard many differing opinions on which is better, RAID 1 or RAID 5. If I have 2 Large disks (say 36GB) on a Raid-1, I assume having 4 smaller 9GB drives on a RAID-5 will be faster, but I am not sure due to overhead and the like.
Does anyone know where I can get more information on RAID performance and how it is going to affect me? The database is going to be Read-Write, with a ton of small transactions, and the occasional (usually on a weekend) aggregation.
can anyone help me? I just installed WinNT 4 SP3 and now trying to install SQL 6.5 on an IBM Netfinity 5500 with RAID 5. There is one 500mb partition for the system and then one 12 gig partition and SQL only seems to see ~117mb of the 12 gig partition. Is this a known prob? Thanks for any input Josh
