Frequently Asked Questions

We welcome your questions, comments and bug reports. If you don't find the answer below, try the Support page or search our Google Group. If you still can't find the answer, contact us through our Google group. We generally respond within 24 hours.

My registration failed, Help!

There are two ways a registration might fail:

1. You registered successfully, but did not receive an activation email. Globally servers are very concerned about SPAM and pfishing, so our activation email occasionally does not make it to the user. If you don't see the email in your personal SPAM catcher, it is best to just contact us, and we will activate on our side.

2. You tried to register, but received a warning that your registration could not be completed. This may happen because CIPRES does not allow a user to hold multiple accounts. The application automatically prevents registration of a second account by a single user. If you see this message, just contact us through the email referenced in the message. We can help.

Where can I learn how to use CIPRES?

Basic CIPRES Science Gateway Usage:
The CIPRES Science Gateway interface manages data in folders, much like an e-mail client. The data is then used to stage individual jobs (also stroed in folders). Here we provide some flash demos to illustrate the basic functionalities of the CIPRES Science Gateway. Each demo is less than 3 minutes.

Known General Issues: here.

Code Specific Issues: here.

Supported Formats: here.

Can you handle GB input files?

CIPRES offers a Globus file transfer tool that allows you to upload files that are in the GB range without using http protocols.
Globus uploads are fast (if your internet has a decent upload speed) and robust/resumable. CIPRES can handle individual files up to 16 GB. You can find more information about Globus transfers on our video support page.

How does CIPRES support its users?

We are committed to providing excellent user support.
Our team is small, so perfection is not possible. But we are eager to learn about and attend to any issues users encounter.
Support is avaialble (minimally) M-F 9 AM - 5 PM Pacific Time. Response time is typically 2-4 hours during that time.

Bugs/Job Issues:
We encourage users to report issues through our Google Group.

Custom help is available:
Our team is highly experienced in running community HPC codes successfully. We are happy to work with individual users to provide advice, make new features and codes available, and to customize runs to meet user needs whenever possible. 
 

Can I monitor results in real time?

Intermediate Results: CIPRES makes it possible to view intermediate results so you can follow the progress of your job. You can (and are encouraged to) monitor the progress of long runs. To view intermediate results for a running job, click on the "View Status" button. When that page opens, you will see an "Intermediate results" hyperlink. Click it and you and a pop-up will appear that displays all input files and any results for the selected job. The intermediate results pane shows input files and scheduling information files immediately, so you can make sure everything looks fine before the job begins. When the job initiates (normally within a few minutes), more files will appear, beginning with start.txt. The start.txt file notes the time when the job began executing. More details are provided here

How do I know if my job has started? If the job has not started, the intermediate results pane shows only input files and scheduling files. This means your job is still waiting in the queue. Jobs usually (but not always) start within 15 minutes. You can keep the intermediate file page open and refresh from time to time to see when the start.txt file appears.

What if the queue time seems long? If the queue time seems long, there are two possible reasons.
First, and most likely, the queue may be paused because of a scheduled or unscheduled maintenance. Typically, when a maintenance occurs, we will post a message on the front page. The (normally green) status box will turn red, with an informative message. You can also check for scheduled maintenance events here; or just contact us at the link below if you aren't sure. Any job that might be still running into the maintenance period will be held in the queue until maintenance is complete. Once maintenance completes, the job should initiate within an hour or so. If it does not, you should contact us at the link below. Occasionally (but rarely) a job does not come out of maintenance cleanly, so it is possible that your job will hang.

Second, and less likely, the queue might actually be long, and it could take hours for the job to start. This is quite rare, however. If you have questions, please contact us.

My job died, now what?

If your job fails immediately, or dies unexpectedly, first, please do not delete the job from your interface.
Try using the help tools to troubleshoot the issue, and if you can't figure it out, report the failure to us. Here is how.

A troubleshooting video is available.

We will happily credit your account for any loss due to an issue on our side.

 

My job completed, but I don't see an output tree

What if the job completes and I don't see my tree?

There are basically two reasons this might happen that are in your control.

The most common reason for a job failure is that the code did not like your input files or your command line choices. To diagnose this, have a look at the last few lines of the stdout.txt and stderr.txt files. Most of the codes we support provde notice of a failure a useful and human readable statement of the failure case. You can ask us for help if you cant make sense of the message.

The second common reason for a file not appearing is the job reached its wall time limit before it completed. If there is no obvious error message in stderr.txt or stdout.txt, look next at the file _scheduler_stderr.txt. If the job did not complete in the allotted time, it will contain a message like this:  slurmstepd: error: *** STEP 23338670.0 ON exp-1-43 CANCELLED AT 2023-06-23T15:01:48 DUE TO TIME LIMIT ***.

To handle this, see "restarting jobs".

If you don't see evidence of either of the above, most likely the job failed due to a transient system problem (this should be rare), or a specific known issue with the code (see the text for specific codes below). Please report any jobs that fail to produce the expected results.

A troubleshooting video is available.

My job didn't complete, can it be restarted?

My job didn't complete, can it be restarted?

Restarting Jobs: If it is supported, you can "restart" a job that "timed out". If a job can be restarted, a "restart" button will appear next to the "clone" button. You may have to "refresh" the page using the button at the top right of the jobs list page. Just click the restart button, and the job will run again for the same amount of time you originally configured. The interface will allow you to review and alter many parameters. The following codes can be restarted on this way: BEAST, BEAST2, MrBayes, IQ-Tree, ExaBayes, IMa3, RAxMLNG, PhyloBayes.

Special Restart Interfaces: You can restart MrBayes using a specific restart interface. These interfaces are more tedious, but will work.

No Restart Option: If your code does not support restarts, your main option is to "clone" the job, and re-run, using the parameter pane to increase the maximum allowed run time.  If you need to run longer than our maximum allowed time, all is not lost, please contact us for help and advice.

How do you handle maintenance?

All complex HPC resources must undergo maintenance from time to time. Our goal at CIPRES is to minimize the risk to your allocation time. Maintenance periods are typically planned and scheduled in advance, but (rarely) an emergency situation arises, and the resource must be shut down unexpectedly. When a maintenance period is required, the queue is paused because the maintenance will kill running jobs. Any job that might be still running into the maintenance period will be held in the queue until maintenance is complete. We notify users of planned maintenance, or ongoing maintenance using a status box on the front page. The (normally green) status box will turn red, and provide an informative message. Please contact us at the link below if you have questions.

During any maintenance period, submissions will be paused (though your job can be saved for later deployment). Once maintenance completes, the job should initiate within an hour or so. If it does not, you should contact us at the link below. In the event of emergency maintence that kills a running job, please contact us to refund time to your account.

If you are running a restartable job, you can work through a coming maintenance period by shortening your maximum run time, so it does not enter the maintenance period. If the run ends before it is complete, you can restart and complete the job once maintenance is over. Just be sure your job is restartable.

What parallel codes does CIPRES offer?

Parallel tools currently supported by the CIPRES Science Gateway can be found here.

*GPUs are charged as 20 cpu cores. So a 4 GPU run will be charged as 4 * 20 = 80 cpu cores per hour.

What serial codes does CIPRES offer?

Serial tools currently supported by the CIPRES Science Gateway. We also support several utilities, such as READSEQ, NCL-Converter for file transformation. All of these codes run on a single core.

Tool Version Language Compiler Cluster
ClustalW 2.1 C/C++ Clang Expanse
Divvier 1.0.1 C/C++ GNU Expanse
MPBOOT 1.2 C/C++ GNU Expanse
PREQUAL 1.02 C/C++ GNU Expanse
PROBCONS 1.12 C/C++ GNU Expanse
PROBALIGN 1.3 C/C++ GNU Expanse

How can I subscribe?

A free trial is provided with your registration. Up 1,000 hours are free for the first six months. If you need more time, our subscription services are available.

Purchase subscription with a credit card.

Contact us to purchase a subscription with a purchase order.

The US National Science Foundation and the US National Institutes of Health have supported CIPRES operations for the past 10 years, but continued CIPRES operations now depend on revenue from subscriptions. The cost of a subscription is modest, and provides:

  • Up to date versions of 20+ carefully benchmarked tools.
  • Up to 200,000 core hours of compute time
  • Automatic run configuration to ensure highly efficient use of precious compute time.
  • CPU or GPU nodes ar used, depending on the code and data set.
  • Up to 200-fold increase in speed relative to single processor runs*.
  • 150 GB of indefinite storage.
  • Robust, resumable transfer of multi-GB files via Globus.
  • Online support when you encounter problems.
  • Purchase from your grant with a procurement card or purchase order.

*Speed-up depends on the individual code and data set.

The subscription levels available and costs are listed below. Subscriptions are guaranteed for one year from the date of purchase.
If you the current offerings don't meet your needs, please contact us.

If you have further questions about subscribing, please see our FAQ page.

Prices in effect for all non-commerical users, beginning Jan 1, 2023. Prices are for a one year subscription by default, contact us if you have other requirements. Commercial users, please contact us for pricing.

Level Compute Time Annual Cost Cost/1000 CPU Hour*
1 1,000 core hrs $30 $30
2 10,000 core hrs $170 $17
3 20,000 core hrs $280 $14
4 50,000 core hrs $500 $10
5 100,000 core hrs $900 $9
6 200,000 core hrs $1800 $9

* Cost per CPU hour is calculated assuming use of all available CPU hours at the subscription level indicated.
.

I have a question about how subscriptions work

CIPRES Subscriptions: Frequently Asked Questions

What does my subscription fee pay for?  A subscription supports ongoing CIPRES operations. The activities funded by subscriptions include maintaining and improving the existing interface, testing, benchmarking, and releasing new tools, providing user support, monitoring and responding to interruptions in service, ensuring users have equal access to compute resources, helping individual users with special job requirements. Subscription proceeds will only be used to defray current costs, so we can continue to provide service to the community. CIPRES will always be a not-for-profit project. The US NSF still provides the compute time at no cost through the ACCESS project, and as a result, the cost of using CIPRES is much less per cpu hour than a commercial cloud provider.

Subscribe now. 

Will I be charged based on how much I use the resource, or is it a flat rate? Our model is based on a flat rate, renewable annual subscription. A Level 1 subscription provides access to CIPRES  codes, job submission tools, 150 GB free storage, user support, and 1,000 core hours of compute time per year. Enhanced subscriptions provide the same benefits and access to more compute time. The subscription levels provide 10,000, 20,000, 50,000, 100,000, or 200,000 core hours of compute time per year.

How will my usage be calculated? Usage is calculated based on the amount of compute time you consume in a calendar year. Your account is charged only for compute time you actually use. Most CIPRES jobs run on multiple processor cores, so your account will be charged for the number of compute cores your job used multiplied by the number of hours your jobs run. Further submissions from your account will be halted automatically when you reach your annual limit.

What happens if I need more time? You can access more compute time by purchasing a new subscription at any time. Any user account can purchase a new subscription at any time. Should you require more than 200,000 hours for a year, please contact us.

What happens if I don’t use all of my available time? By default, compute time available at the end of subscription year will not be carried over to the subsequent year, the year begins on the date of the most recent purchase. So you can extend the availability by making an additional purchase.

Can I charge this to my research grant? Yes. CIPRES will accept institutional procurement credit cards as well as purchase orders. Payment must be received before access to resources will be provided.

Why shouldn’t I just get my own ACCESS account and run jobs from the command line there? This is possible if you are at a US institution. However, ACCESS accounts can only be requested by users in the US who have a Ph.D. or are at the Principal investigator level or higher in their organization. ACCESS accounts do not provide a user interface, but instead rely on command line usage. The user is responsible for installing all of the codes they require, and configuring them to run efficiently. CIPRES handles all this for you.

Why shouldn’t I just use a cloud provider instead? It is possible to run your jobs through a cloud provider. However, this will require learning to use the command line interface, installing the tools you require, and benchmarking them so they run efficiently. The cost of running one CPU (provisioned like ours) at AWS is  8.5 cents (US) per hour, and this does not include network and data storage charges. In this model, 5,000 core hours on a “bare metal” AWS instance will cost at least $425 plus the amount of time required to set up the instance and install the code. For $150, a subscription at CIPRES provides 11,000 hours of compute time, access to current tools, 150 GB data storage, a user-friendly point and click interface, and customer support.

Where can I get the CIPRES Software?

The software for the CIPRES Workbench framework is freely available.
Instructions can be found here.
The paper describing the software is available here

How do I cite CIPRES

If you use the resources available from the CIPRES Science Gateway to complete published work, please cite us as follows: Miller, M.A., Pfeiffer, W., and Schwartz, T. (2010) "Creating the CIPRES Science Gateway for inference of large phylogenetic trees" in Proceedings of the Gateway Computing Environments Workshop (GCE), 14 Nov. 2010, New Orleans, LA pp 1 - 8. DOI: 10.1109/GCE.2010.5676129 PDF.

hummingbird in flight

Get 1000 Hours free

On the UCSD Supercomputer

Start Your Trial