| Commit message (Collapse) | Author | Age | |
|---|---|---|---|
| * | Use with-store/non-blocking from guix | Christopher Baines | 2024-05-14 |
| | | | | | | As it opens a non-blocking connection, rather than doing a blocking handshake then making the socket non-blocking. | ||
| * | Remove with-timeout | Christopher Baines | 2024-05-10 |
| | | | | | As it's now unused. | ||
| * | Avoid with-store/non-blocking | Christopher Baines | 2024-05-10 |
| | | | | | | As I'm seeing exceptions around here, but I think the syntax rule may be hiding exactly where. | ||
| * | Have the port monitoring fiber use non blocking sockets | Christopher Baines | 2024-04-28 |
| | | |||
| * | Don't use open-connection-for-uri from Guix | Christopher Baines | 2024-04-28 |
| | | | | | As it's timeout approach doesn't work with suspendable ports. | ||
| * | Use the new process metrics exporter | Christopher Baines | 2024-04-17 |
| | | |||
| * | Fix the sleep-impl argument to retry-on-error | Christopher Baines | 2024-03-16 |
| | | |||
| * | Add a definition of retry-on-error which uses the fibers sleep | Christopher Baines | 2024-03-15 |
| | | |||
| * | Remove the core-guile-sleep variable | Christopher Baines | 2024-03-15 |
| | | | | | As this is unnecessary. | ||
| * | Actually use non-blocking ports for network requests | Christopher Baines | 2024-03-02 |
| | | | | | In most places at least. | ||
| * | Change retry-on-error to take #:ignore and #:no-retry | Christopher Baines | 2024-01-12 |
| | | | | | And change #:ignore to better reflect ignoring the exception. | ||
| * | Improve thread pool and work queue backtraces | Christopher Baines | 2023-12-09 |
| | | | | | To try and get more information about "conversion to port encoding failed". | ||
| * | Guard against exceptions in the thread pool | Christopher Baines | 2023-12-06 |
| | | |||
| * | Don't check too-many-threads? in the thread pool for each job | Christopher Baines | 2023-12-06 |
| | | | | | | As this is excessive. Instead just check this when there are no jobs available. | ||
| * | Move a hash-set! in the thread pool | Christopher Baines | 2023-12-06 |
| | | | | | To avoid unlocking and locking the mutex. | ||
| * | Re-fetch running jobs in the thread pool | Christopher Baines | 2023-12-06 |
| | | | | | As the information might be stale when wait-condition-variable returns. | ||
| * | Don't pass time to wait-condition-variable | Christopher Baines | 2023-12-02 |
| | | | | | | | | | | | My intent here was to avoid threads waiting indefinitely for a signal. Unfortunately, it seems that specifying a timeout can result in wait-condition-variable entering an endless loop internally (#67194). Hopefully avoid this by not passing a time to wait-condition-variable. Maybe some new threads/fibers will need adding that periodically signal these condition variables to replicate the timeout behaviour. | ||
| * | Remove set-store-connection-timeout | Christopher Baines | 2023-11-20 |
| | | | | | As non blocking connections are being used now. | ||
| * | Avoid readlink syscalls when reading derivations | Christopher Baines | 2023-11-17 |
| | | | | | | | | | | strace -c reports over 10,000 readlink syscalls when reading the derivation for the hello package. By just setting the %file-port-name-canonicalization fluid, this drops to less than 10. I'm not sure if this actually improves performance, but doing less is surely better. | ||
| * | Call wait-condition-variable with longer timeouts | Christopher Baines | 2023-11-15 |
| | | | | | | | The hook processing threads seem to be getting stuck in wait-condition-variable where the timeout is in the past. Maybe setting a longer timeout will help avoid this. | ||
| * | Remove some caching from reading derivations | Christopher Baines | 2023-11-11 |
| | | | | | As I'm not sure this is helpful. | ||
| * | Don't output the whole log for substituting derivations | Christopher Baines | 2023-08-31 |
| | | | | | Just the last bit, as it could be very long. | ||
| * | Fix some logging messages | Christopher Baines | 2023-08-29 |
| | | |||
| * | Pass the store in to substitute-derivation | Christopher Baines | 2023-08-20 |
| | | | | | To avoid an additional store connection. | ||
| * | Include metric for file descriptors | Christopher Baines | 2023-08-20 |
| | | |||
| * | Simplify substitute-derivation | Christopher Baines | 2023-08-10 |
| | | | | | Assume ensure-path is available. | ||
| * | Name more threads | Christopher Baines | 2023-08-09 |
| | | | | | To help with debugging. | ||
| * | Don't use threads for substituting derivations | Christopher Baines | 2023-08-09 |
| | | | | | Instead use non-blocking store connections like Cuirass does. | ||
| * | Fix read-derivation getting the filename | Christopher Baines | 2023-08-03 |
| | | |||
| * | Keep less derivation files open when reading derivations | Christopher Baines | 2023-08-03 |
| | | | | | | This probably isn't the main problem with having too many open files, but it might help avoid bursts of open files. | ||
| * | Make client requests mostly suspendable | Christopher Baines | 2023-07-24 |
| | | |||
| * | Remove unused make-base64-output-port | Christopher Baines | 2023-05-20 |
| | | |||
| * | Improve work queue and thread pool exception logging | Christopher Baines | 2023-05-17 |
| | | |||
| * | Add more information to job raised exception logs | Christopher Baines | 2023-05-14 |
| | | |||
| * | Add #:streaming? to call-with-streaming-http-request | Christopher Baines | 2023-05-11 |
| | | |||
| * | Add missing argument to port timeout errors | Christopher Baines | 2023-05-11 |
| | | |||
| * | Correct some indentation | Christopher Baines | 2023-05-10 |
| | | |||
| * | Separate read and write timeout exceptions | Christopher Baines | 2023-05-09 |
| | | | | | So it's clearer what has occurred. | ||
| * | Remove make-chunked-output-port* as it's now unused | Christopher Baines | 2023-05-08 |
| | | |||
| * | Force sending the request in call-with-streaming-http-request | Christopher Baines | 2023-05-08 |
| | | |||
| * | Stop using chunked transfers for file uploads | Christopher Baines | 2023-05-08 |
| | | | | | | As the amount of data to upload is known, this is unnecessary complexity and overhead. | ||
| * | Add utility to get system uptime | Christopher Baines | 2023-05-05 |
| | | |||
| * | Remove the gbc prefix from the thread names | Christopher Baines | 2023-05-02 |
| | | | | | As this shouldn't be needed to help identify them. | ||
| * | Neaten up the make-chunked-output-port* implementation | Christopher Baines | 2023-04-30 |
| | | | | | | As the comment says, this needs a bit more looking at to determine what (if any) changes should be suggested for Guile. | ||
| * | Stop monitoring uploads through the chunked output port | Christopher Baines | 2023-04-30 |
| | | | | | Use the dump-port* progress reporter instead. | ||
| * | Try to instrument ports/file descriptors | Christopher Baines | 2023-04-27 |
| | | | | | | | I'm seeing "too many open file" errors, and while I'm not sure if port-for-each is directly connected, it sounds like it might be worth instrumenting. | ||
| * | Remove chunked input code | Christopher Baines | 2023-04-25 |
| | | | | | This is now upstream in Guile. | ||
| * | Remove some thread pool logging | Christopher Baines | 2023-04-22 |
| | | | | | As I'm not currently debugging this. | ||
| * | Use set-car! when sorting the prioritised work queue | Christopher Baines | 2023-04-20 |
| | | | | | | As stable-sort! probably doesn't guarantee to alter the list as I was previously expecting. | ||
| * | Attempt to make with-port-timeouts handle the GC restarting poll | Christopher Baines | 2023-04-17 |
| | | | | | | | From what I'm seeing in strace, I think the GC is breaking the timeout behaviour by restarting the syscall, these changes should work around that behaviour. | ||