Beep beep's questions

My company is trying to figure out what type of SAN to purchase. This is specifically for database servers that are becoming IO constrained (storage is DAS right now, but we're hitting the limit of a single server and we'd like to add clustering as well).

We need a solution that will produce around 3000 IOPS long-term (we currently peak around 1000 IOPS). Most of our database operations are small reads/writes. Based on discussions with HP engineers and others online, an HP P2000 with 24 SAS HD's in a RAID 10 configuration will deliver just short of that speed for ~$20K. Add in controllers and other items to build out the SAN puts us right around our max budget of $30K.

But online, I see that many SAS SSD's deliver speeds of 80,000 IOPS+. Is this realistic to expect? If so, would it be realistic to get a P2000 or similar entry level SAN and throw a few SSD's in there? Our databases are small, only a couple TB total. If we did this, we'd have the money leftover to buy a second SAN for mirroring/failover, which seems prudent.

I have DTC enabled on 2 servers - a web server, and a database server. There are a few places in our web applications where we need to use DTC (via a TransactionScope) to manage transactions involving multiple databases. On our AD servers it is working fine, but for some reason it isn't completely working on one pair that are not members of a domain. I get the following error:

 MSDTC on server 'DATABASE-SERVER' is unavailable.

Using DTCPing from my web server, I get a successful test, but from my database server I do not (it stops after the RPC call with no messages).

Some notes on configuration:

• Firewalls are disabled on both servers
• DTC is configured the same on both (the Local DTC Properties have Network DTC Access et, allow remote clients/admin, allow inbound/outbound communication, enables all transaction types, and 'No Authentication Required'
• Using DTCPing, I can connect from the web server to the database server, but not vice versa
• RCPPing works fine in both directions
• DTC is running on both
• I've restarted the service, and performed an uninstall/install multiple times on both servers
• nbtstat -c correctly shows the remote server names

The only difference between the servers I can find is that when I initiate DTCPing from the database server, it doesn't finish the test (it just says "Please start PING from WEB-SERVER to complete the test"), while if I initiate from the web server it complete the entire thing on both ends.

We have a web application that utilizes SQL Server 2008 on a single database server. All storage is local. For the past year we have been trying to get any form of SQL Server Replication to work with our configuration, but it won't. The reason is because we have over 2,000 databases that are constantly being updated (one for each of our clients), so our tests show that all forms of replication are too resource intensive.

Every time I ask this question, people focus on the fact that we have too many databases. This is something that cannot change (for regulatory and other reasons), so I'd like to focus on how we can replicate the data.

We've been told that one option is to move all the data to a SAN and let the SAN replicate the data (or take frequent snapshots). However, if our database server fails, is there a risk of a database getting corrupt in this case? Is it possible to leverage the SAN, replicated to another SAN, to provide a decent DR solution (in our case, we can lose up to about 30 minutes of data, but we cannot lose an entire day's worth ... i.e. we can't go to a prior night's backup).

Our company is having a real problem with spam, phishing, and sophisticated viruses (ones that are brand new at time of first download, and not recognized by any virus scanners for at least a few hours after being downloaded, sometimes days). We have needed to wipe a few machines as a result, users have been caught by phishing scams, and we even had one virus that captured some of our info from a file share.

I am wondering how other companies guard against these types of threats. We've tried educating users about what to click or not click, but no amount of education seems to eliminate the problem (particularly for non-tech users). Do companies set up secure browsing/e-mail environments (e.g. as a separate Virtual Machine)?

FWIW, we're running an Astaro Firewall and have two anti-virus programs (ESET and TrendMicro)

Are there any significant drawbacks to joining a SQL Server machine to the domain (i.e. security, performance, etc)? Our box is currently off the domain so we just manage users locally, but we were thinking that security may be a little stronger with it off the domain.

Edit - Forgot to mention, this is a database server for a public-facing web app ... not sure if this changes the answer (the web app is in its own DMZ).

We are about to release a web application for our users, and are trying to figure out if others put their servers in a DMZ, or just keep it off the domain behind the firewall and greatly restrict access via firewall rules? Has anyone here found any cons to just restricting access via firewall rules and O/S perms?

Note that this site is an ASP.NET MVC front-end with SQL Server back-end.

Also - this is an HR/Finance application, and the database back-end contains our most valuable data. From a security perspective, I'd rather give the world root access to intranet than allow access to the database server. As a result, my original plan was to skip using a DMZ and only opening up port 443 on the firewall to the web server ... this seemed no worse than putting the db server on the DMZ with the web server.

We are trying to deploy Google Chrome across our corporate network, but we're finding that it takes 2-4X longer to load an https page (particularly our own internal ones) compared to IE. Has anyone experienced this and found a fix?

Update

Based on Handyman5's suggestion, I ran some diagnostics in Chrome, and found that the largest amount of time (over 90% on each page) was being spent on pulling static files from Cache and rendering the page. However, if I turn off SSL on our site, this is almost instantaneous.

Any thoughts on why this would be?

This seems really basic, but I couldn't find an answer already ... we are looking to deploy a website to an IIS server, and a database to a separate server. We aren't sure how to best secure this for access by internal users and internet.

The IIS Server is not connected to the domain, and neither is the database at the moment. Should it be? Also, if we use SQL Server authentication instead of Windows Authentication, should we just hard-code the user/password into our internal apps?

We are building an application where each client will have their own database. None of the databases are particularly large (20MB to 400MB each), but there will be ~5,000 to start and at any one time 100 or so will be active.

Our team has been debating how best to set up the system. Clients only access their database once every 2 weeks (401k/finance processing), and only use it for 10-30 minutes at a time. Operations are evenly spread between reads/writes.

Half of our team feels that as a result we should spread the databases across multiple cheap servers and just use SQL Express ... they say that memory/caching wouldn't be as useful given the short period of time each database is used (we don't have the budget for the full SQL Standard on more than 1 server).

Is this the case? A higher memory limit is really the only advantage I see MSSQL Standard bringing us (we already have scripts for doing backups/restores, schema upgrades, migrating data, etc).

Update

I'm particularly interested in performance characteristics of multiple databases vs. one database. Wouldn't the end user experience be better hitting a single 200MB database than a 1TB database (even if both were well indexed)? This also means we can easily backup/restore single client databases very very fast, right? Would we need to tune SQL Server to better handle the 'thousands of databases' scenario?

A company I sometimes consult for is relatively small (40 employees), and they have a royal mess of Office licenses. OEM, Retail, Upgrades, O2K/XP/2K3/2K7, etc. They basically buy whatever retail license they can find cheapest online, and have someone track it all in a spreadsheet for compliance purposes. They also use a Microsoft Action pack license for getting another 10 copies of Office/Vista for free.

While it all seems to follow Microsoft's licensing rules, it also seems horribly inefficient. I've talked to them about Microsoft's Open license, but they don't see any advantage to it. What do other relatively small businesses do? Are Open licenses popular, or do most of them just buy retail like my client?

We have a network of 16 heavily used Windows 98 PCs connecting to a Server 2003 machine. These machines are experiences long delays when opening/running files on the network during times of peak usage (1:30-2:30pm daily), sometimes lasting a minute. The user will just get a black DOS screen as the executable is waiting to open, where usually it would pop up in well <1 second. This is particularly vexing because Server 2003 performance counters do not show a long period of high load - most stats never even reach above 5%.

When these 16 clients were connecting to a 6 year old DL380 running Linux, we had some slowness, but no delays. We recently transitioned to Active Directory, and upgraded the file server as well. The file server has dual quad core processors, 4X more memory, more disks, faster drives (15K vs. 10K), etc ... so it doesn't seem to make sense that it can no longer handle concurrent processing.

Since we're using Win98, is there some setting that we should change on 2003 to optimize disk read/writes? All file operations are to flat files, and all writes are done by a single user (multiple users cannot access the same database at a time). Any other thoughts on how to track this down ... we haven't had any luck over the past month.

Note: unfortunately we cannot upgrade beyond Win98 for another 2 years, we're waiting for legacy software to be rewritten and this critical software only runs on Win98 due to use of hardware interrupts.

We recently upgraded from an old Linux server running Samba to Server 2003. All of our clients are Windows 98 virtual machines, and heavily access the server for file services. ~95% of the time, there are no problems and all runs smoothly ... but at least once per day per machine, and sometimes much more frequently, the network shares become inaccessible from a program until the user clicks on the drive through explorer (which sometimes takes 10-20 seconds to access when the system is in this state). The problem is that soon after connectivity to the shares is lost, our legacy 16bit app typically freezes and locks Win98 ... which also corrupts data.

We've already disabled the autodisconnect feature on Server 2003, Win98 has all the patches, and the newest DSClient (with patches) is applied on Win98.

The only universal symptom we've noticed is that when a network share becomes inaccessible/frozen from DOS the shared drive shows up as 1.99GB instead of 135GB in explorer. For example:

Both servers have 135GB RAID arrays, and typically show up as 135GB. The moment the network share become patchy, we took this screen shot. This symptom seems to always happen right before the Win98 app comes down, can happen anywhere from 2 minutes to 8 hours after logging in, and occurs on all 38 of our Win98 VMs (some of which were built differently or use a different VM player just to see if that was the problem).

Any ideas?

BTW - We cannot move to 2000/XP/etc for this legacy app because it is DOS-based and uses printer interrupts and such. We're in the process of writing a replacement, but that won't be ready for another 10-11 months.

We are in the process of building a new secure web application for our company, and are looking at options for ensuring redundancy in our data center. We don't have a lot of money, but have relatively high bandwidth requirements, so have considered a 50/20 FiOS Business connection with a T1 backup, which would cost ~$600/month total. In comparison, a 15Mbps partial T3 costs $2500/month in our area.

If we were to go with this set up, is there a way for someone typing in our address to automatically be redirected to the other line if the FiOS is down? Since it's an SSL connection, we'd need to ensure that the certificate would work either way as well.

Update - to those suggesting an outside data center solution, we have relatively high internal user bandwidth requirements (~80Mbps). Purchasing an 80Mbps pipe to a colocation facility, to which we'd also have to pay high monthly fees, isn't really financially feasible. In our area, such a setup would cost around $10,000/month.

We currently have five 24 port 3com unmanaged switches. Four of them are 100Mb switches with no gigabit port (MDI Uplink port is 100Mb). The fifth is a full gigabit switch which we use for all of our servers. They are daisy chained together at the moment, kind of like this:

At certain times of the day, network performance becomes atrocious. We've verified that this has nothing to do with server capacity, so we're left with considering network I/O bottlenecks. We were hoping to connect each 100Mb switch directly to the gigabit switch so that each user would only be 1 switch away from the server ... like this:

The moment this was connected, all network traffic stopped. Is this the wrong way to do it? We verified that no loops exist, and verified that we didn't have any crossover cables (all switches have Auto-MDI). Power cycling the 5 switches didn't do anything either.

We are about to roll out a dual web/internal transactional application where each client has their own database. Each database is very small - under 50MB each, so we were wondering if it would make sense to use SQL Express 2008 instead of the full SQL Server.

This seems to have the advantages of distributing disk I/O across servers while saving massive $$$ (since small 15K drives and used dual core servers are both inexpensive). If at some point we need too many servers, we can upgrade to SQL Server ... but with dozens of internal users this just seems too expensive right now (particularly since we'd need a failover box).

1GB memory and usage of 4 cores on a single processor doesn't sound too restrictive given our small database sizes. We'll never have more than ~200 concurrent users, and most operations will be more transactional (which seems to favor lots of high speed disks over heavy RAM/CPU, right?)

Am I missing any advantages of SQL Server Standard that could justify the extra $5-20K investment initially?

We currently have two 6 year old linux file servers (an HP DL320 and DL380) that are starting to be dogs. The DL320 just uses 10K IDE drives set up in RAID1, the 360 has 6 drives set up in RAID10. File services are via Samba, and the DL320 is having far more problems at the moment.

We intend to move to Windows Server 2003 (2008 won't support our old Win98 workstations - we can't get rid of them for 3 more years when our legacy software is converted). However, I'm wondering if it makes more sense to buy a SAN instead of buying a newer machine with faster drives?

Our specs

• 40 users
• Linux-based servers, migrating to Windows Server 2003
• Linux 'top' shows samba often in a wait state (high load)
• Linux 'sar' command shows many small transfers, but only moderate bytes/sec (so not a network bandwidth issue)
• All users connecting directly to file server with old chatty DOS client application (not client/server), which results in thousands of tiny transfers per minute

Any thoughts?

Possible Duplicate:
Can you help me with my software licensing question?

If we are buying new or used Windows FPP (Retail), we make sure we keep the Media, EULA, Manual, and License. Microsoft had informed us that all of these materials were required for proper license compliance. However, what needs to be kept for an upgrade?

I ask because we are constantly adding new Win98 copies as Virtual Machines to our XP network (our old software requires it ... and yes, we're in the process of rewriting that software). It's becoming VERY difficult to find FPP Win98 SE, so we're considering just purchasing a bunch of Win 3.1 FPP (which are easier to find for some reason) and buying used Win98 SE Upgrades. Many of these upgrades just come with a disk and product key ... Microsoft won't tell us if this is sufficient for compliance since it's dealing with Win98, or if we need to find copies that come with the other material.

Any thoughts on What constitutes a "legal" or "in compliance" upgrade? Thanks.

We are setting up our first Windows Server 2008 web server (previously we used Linux/Apache for all web servers), and I was hoping to get some helpful tips for setting it up and securing our installation.

We will have one web server and one database server (SQL Server 2008) to start, running IIS 7 (obviously I guess). All web pages will be SSL encrypted (i.e. https://).

We have a subversion repository behind our firewall, and I'm trying to figure out how to provide remote developers with access (without giving them access to the whole machine or VPN).

Is there a way that we can have our webserver point to the subversion repository, and redirect all subversion actions? This way the developer only needs to access our web server, not the secure file server which also holds the subversion repository.

My company has ~40 employees, all using POP3 mail accounts on a shared host. They allow up to 200 email boxes, have no limit on the # of outgoing/incoming e-mails/hour, and we've never had really big problems with them (plus, they're REALLY cheap).

However, we suspect that at some point before we reach the 200 email limit we'll need to upgrade ... but this is honestly just a baseless suspicion due to so many other companies our size using Exchange or other mail servers. My question - how should we guage the right time to upgrade, and what should be our criteria for the next solution?