Dentrix, Eaglesoft and Open Dental: what to check before you call anyone
It is twenty to nine, the first patient is already in the chair, and the schedule will not load. This guide covers the two minutes before you call anyone: the one question that tells you how bad it is, the things anyone at the office can safely check, and the short list of obvious fixes that will cost you the patient database if you try them.
It is written for the person at the front desk, not for an IT department, and it names what actually breaks in Dentrix, Eaglesoft and Open Dental specifically, because the three fail in different ways and the habits do not carry across.
First question: one machine, or all of them?
Before anything else, walk to a second operatory and try there. The answer sorts the morning into one of two very different situations, and it takes about thirty seconds.
If it is one machine, the fault is on that machine. The rest of the practice can keep working, the patient in the chair is fine, and this is a today problem rather than a right-now problem. Move the provider to a working operatory and carry on.
If it is every machine, the problem is the server or the network between it and the operatories, and the practice is stopped until it is fixed. Make that call immediately, before trying anything else.
It is also the first question a provider who works in dental practices will ask you, so having the answer ready removes a round trip. A provider who does not ask it is telling you something too.
What anyone at the office can safely check
All five of these are observational or reversible. None of them can make the situation worse, and any one of them can save you an hour of somebody else's time on the phone.
- Is the server actually on? Look at it. Do not assume. A server that has powered off, gone to sleep, or is sitting on a restart-required screen after an overnight update presents exactly like a network failure from the front desk.
- Is the server plugged into a cable? Patterson requires that an Eaglesoft server be wired to the network and says it cannot use a wireless connection. If somebody has moved the machine and it is now on Wi-Fi, that alone explains intermittent failures across the whole office, whichever of the three you run.
- How much free space is on the server drive? A full disk stops a database dead, and it is the most common preventable cause of a lost morning in any of the three. It is also the one that gives the most warning, since it arrives over weeks.
- Did anything update last night? Imaging software and the practice management database want the same resources at the same moment, and an imaging update that shipped without anyone testing it against the other one is a classic Wednesday morning. So is a Windows update that installed and is waiting on a reboot.
- Restart the workstation. Not the server. Restarting the machine in front of you is safe and fixes a real share of single-machine faults. The server is a different object with a different rule.
The four things not to do while patients are in the building
These are the instincts that feel most helpful and do the most damage.
- Do not power-cycle the server. The patient database is held open by a service that runs continuously the whole time anyone is in the software. Pulling power on that is an unclean shutdown of a live database. It often survives. When it does not, the recovery is a restore, and the restore is only as good as the last backup anyone actually tested.
- Do not tidy up the data folder to free space. Dentrix keeps an exported copy in a folder named for backups, Eaglesoft keeps its database and log file together under its own program folder, and Open Dental keeps a MySQL data directory. Those files are the patient record, and the file that looks most like a disposable log is frequently the one the database cannot open without.
- Do not uninstall or repair-install to clear a hang. It rarely addresses the cause, and it converts a problem somebody could have diagnosed in twenty minutes into a reinstall with licensing and version questions attached, in the middle of a working day.
- Do not restore last night's backup to get back to normal. Restoring is not undo. It overwrites whatever is currently there, including anything charted this morning, and it assumes the backup is restorable. Open Dental's own documentation is direct that its built-in tool can be used for backups only and not for restores.
What actually breaks, by product
The three products are not variations on a theme. They store data differently, they fail differently, and a habit carried from one to another is how a practice loses a database. Everything below comes from each vendor's own published documentation.
Dentrix
A service runs on the server at all times and holds the database open for every workstation. When that service is stopped, every operatory loses the software simultaneously, which reads at the front desk as a network outage. Restarting it is often the entire fix. It is also a server-side action, so it is not a front-desk job.
Because the database is permanently locked, Dentrix exports a separate copy for backup software to collect. That is a good design with one catch: if the export stops running, your backup keeps succeeding against an increasingly old copy and nothing reports a problem.
Eaglesoft
Eaglesoft runs a SQL Anywhere database engine, and unlike Dentrix there is no live export. The engine has to be stopped for a clean backup, so the backup window is real downtime the practice has to agree to. It does not happen invisibly at three in the morning.
Patterson documents a limitation that matters more here than anywhere else, and they do not hedge it. Under Windows Server 2019, 2022 and 2025 alike, their requirements list one line: scheduled tasks fail to start and stop the Eaglesoft database. Windows Server 2016 is the only one without it. So the ordinary way of automating an Eaglesoft backup does not reliably work on any current server operating system, and the office finds out at eight in the morning.
The same document lists McAfee Virus Scan as not compatible with Eaglesoft and Windows Defender as supported. It also says the server must be wired and that Eaglesoft is not supported with any WAN solution. So McAfee on an Eaglesoft server is a documented problem, not a matter of opinion. You can read it yourself in Patterson's Eaglesoft hardware and network requirements, March 2026 revision.
Open Dental
Open Dental is a MySQL database, which makes it the most open of the three and the most dependent on how it was set up. Two things break practices repeatedly.
The first is a connection ceiling. Open Dental notes that a server running Windows Professional rather than a server operating system is limited to twenty concurrent connections. A practice that was fine for years hits it after hiring, and the symptom is that the software works first thing and starts refusing people mid-morning as more machines and devices connect. It looks like a performance problem and it is a licensing ceiling.
The second is backup software that does not understand a live database. Open Dental names consumer file-sync and backup products, Carbonite and iDrive among them, as having corrupted databases. Their documentation also states McAfee must be uninstalled from the server, which is the second of the three vendors to say so about the same product.
The error numbers, and what they are really telling you
Most of the numbers that come up are not dental software errors at all. They are Windows errors surfacing through whichever program happened to ask first, which is why searching the number alone returns results about accounting software and games.
- 0x800706BA, the RPC server is unavailable. A Windows networking condition. One machine tried to reach a service on another and got nothing back. It often follows a Windows update, a computer being renamed, or a change to the firewall or antivirus. Restarting the workstation clears a fair number of these, and that is safe to try.
- 1310, error writing to file. A Windows Installer permissions problem. Something does not have the rights to write where it is trying to write.
- 1935, an error during the installation of an assembly. Also Windows Installer, and usually a damaged .NET or Visual C++ component underneath rather than anything to do with dentistry.
The useful part is not the definitions, it is what the number tells you about timing. 1310 and 1935 only appear while something is installing or upgrading. If you are seeing one of those, somebody is running an update, and the answer is to stop and finish it properly rather than to keep clicking. An update is not a thing to be doing with patients in chairs.
A number that appears while the office is simply working is the other kind. It points at the network or the server, it belongs to the first question in this guide, and one machine or all of them still decides how urgent it is.
One thing that is not a crash, and is worth asking about anyway
Open Dental leaves the database password to the practice. Their MySQL security page says it plainly: for small offices on small closed networks, it is up to the practice to decide whether or not to set MySQL passwords, and they recommend that it is done. A default installation ships with a user account that will accept a connection from any machine on the network.
Neither Dentrix nor Eaglesoft leaves that decision to the customer, so it is worth one question at your next conversation with whoever supports you: is the database itself password protected, or does being on our network get you in? Reaching the server is not the same as being let in, and on an inherited practice the answer is often no.
Ask the question. Do not fix it yourself. Changing that password without updating every workstation at the same time stops the entire practice.
What your software vendor's support actually covers
Practices routinely assume the software subscription covers more than it does, usually at the moment they need it to.
- Dentrix supports the current version and one version prior, and requires a current service plan.
- Eaglesoft support covers version 19 and above.
- Open Dental is a per-location monthly subscription, and a support call covers up to two specific issues.
None of the three owns your backup. All of them will help with their application. Not one is responsible for the server it runs on, the drive it sits on, the antivirus fighting with it, or whether the copy taken last night can be restored. That is not a hole in the products. It is a job somebody at your practice has to be named for, and at a lot of practices nobody has been.
When it stops being a fluke
Once is a bad morning. Three times in a quarter is a pattern, and a pattern has a cause that is still sitting there between episodes. Failing drives, storage filling up and services restarting are all visible weeks ahead to anyone watching for them, which is the whole difference between finding it on a Tuesday evening and finding it on a Wednesday at nine.
If you want to put a number on what those mornings have cost, the downtime arithmetic uses your own production figures rather than a national average, and it is the number that makes any IT quote legible.
FAQ
Should I just reboot the server?
Not while the practice is open, and not as a first move. The patient database is held open by a service that is running the whole time anyone is in the software, so a power cycle in the middle of the day is an unclean shutdown of a live database. Sometimes it comes back fine. Sometimes it comes back needing a repair you cannot do from the front desk. Reboot the affected workstation instead, which is safe, and leave the server to somebody who can close the database first.
One computer is slow but the others are fine. Is that the server?
Almost certainly not. If the fault only shows on one machine, the fault is on that machine, and the practice can keep working from the other operatories while it waits. That single question, one machine or all of them, is the difference between an inconvenience and an emergency, and it is the first thing a provider who works in dental practices will ask you.
Our backup runs every night and reports success. Are we covered?
A backup that has never been restored is a claim, not a protection. The failure mode is the same in all three products and silent in all three: the job reports success and the database will not open. Ask your provider when they last restored your database somewhere other than your server and what they found. If the answer is that the job has been green for two years, nobody has tested anything.
Would upgrading the practice management software stop the crashes?
Rarely, because the software is usually not what is broken. Recurring faults during patient hours tend to trace to the machine underneath: a drive close to full, disk latency, imaging competing with the patient database for the same resources, hardware past about five years, or an operating system nobody is patching. An upgrade costs money and a day of disruption and leaves all five of those exactly where they were.
What does error 0x800706BA mean in Eaglesoft?
It is a Windows message, not an Eaglesoft one: the RPC server is unavailable. One machine tried to reach a service on another and got nothing back. It commonly follows a Windows update, a computer being renamed, or a change to the firewall or antivirus. Restarting the affected workstation is safe and clears a fair number of them. If it comes back, or if every workstation shows it, the problem is the server or the network between them and it is time to call.
Who do I call first, the software vendor or the IT company?
If the software throws its own error message and everything else on the machine is healthy, start with the vendor. If the whole office lost it at once, if the server is involved, or if it is a question about the backup, start with IT. Worth knowing before you are on hold: none of the three vendors is responsible for the server their software runs on, and none of them owns your backup.
Not sure what your practice is exposed to? The free Security Check is a one-page report on what your office shows the outside world, prepared without installing anything or touching a system you own. The practices page covers how an engagement starts, and if the honest conclusion is that the problem is your current provider rather than your hardware, switching without losing a day is the one to read next.
About the author
Mario Del Mazo is the owner of IT First Response, a managed IT and security practice in Fort Myers serving Lee County, Florida. His background is corporate IT for a healthcare organization operating under HIPAA and HITRUST standards. He publishes his prices, and every client environment is documented so nothing has to be explained twice. More about Mario.