edit-icon download-icon

Common errors of Log Service

Last Updated: Jul 23, 2018

Errors of Log Service

  1. illegal param! [LogContent] is null
  2. send data fail, error_code:WriteQuotaExceed error_message:Write quota exceed projectName:project_name
  3. WARN Aliyun.SLS.Logtail.LogFileReader - logfile xxx.log has n old enough logs with first fail case xxx
  4. Check if the IP address is correct. Currently, Log Service only supports ECS instances in your region
  5. Different operating systems or invalid IP address. Currently, Log Service only supports ECS instances in your region
  6. ShardWriteQuotaExceed

Solutions

1. illegal param! [LogContent] is null

Check the following configurations:

  1. Check if you have entered the log sample.

  2. Check if the regular expression of the first log line is correct.

If your issue persists, provide your log sample and regular expression for us to troubleshoot the issue.

If your issue is still not solved, contact our after-sales technical support.

2. send data fail, error_code:WriteQuotaExceed error_message:Write quota exceed projectName:project_name

If you find an error similar to the following one in ilogtail.log when using Log Service:

send data fail, error_code:WriteQuotaExceed error_message:Write quota exceed projectName:project_name

It means your write quota is insufficient, that is, the log volume you write exceeds the specified threshold or the write speed exceeds the limit.

Currently, the processing capability of each shard is:

  • Write: 5 MB/s, 500 times/s

  • Read: 10 MB/s, 100 times/s

If your data volume is beyond what a shard can process, you can split the shard. For more information, see Split a shard in Manage a shard.

The maximum number of write requests per minute at the project level is 300,000. If you are writing logs by using a program and some requests exceed the quota, we recommend that you write logs in batches or use Producer-lib to limit the maximum packet size and number of logs for each upload to 3 MB and 4096 respectively.

3. WARN Aliyun.SLS.Logtail.LogFileReader - logfile xxx.log has n old enough logs with first fail case xxx

You may have the following error in the ilogtail.log:

WARN Aliyun.SLS.Logtail.LogFileReader - logfile xxx.log has n old enough logs with first fail case xxx

This error is caused by old logs.

Log file collection rules for Logtail:

  1. Process the historical data separately.

  2. Reduce the cache time for data flushing into the disk or even perform real-time flushing.

  3. Note the time zone when logs are modified.

Solutions:

  1. If a new log file is monitored by Logtail, logs in the log file whose log time is not within the last minute are considered as old data and discarded.

  2. For logs newly written to a log file, logs whose log time is not within the last five minutes are also considered as old data and discarded. This error is caused basically because your logs cached in the memory have timed out when they are written to the log file.

  3. Logs whose log time is beyond the range of -7 days–360s are discarded by Log Service. This error is caused basically because the configured time zone of the log is incorrect.

If rule 1 is violated, the historical data is discarded. But if rule 2 and 3 are not violated, no error is reported in the subsequent logs.

If rule 2 is violated, an error is reported occasionally, and you can query some logs in the console.

If rule 3 is violated, some logs are not collected and then cannot be queried in the console.

If the issue persists, contact our after-sales technical support.

4. Check if the IP address is correct. Currently, Log Service only supports ECS instances in your region

An error occurs when you add machines to a machine group in the console, asking you to check if the IP address is correct and telling you that Log Service only supports ECS instances in your region. This error is caused because the intranet IP address of the server you entered when adding the machine group is incorrect.

Check the following configurations if this error occurs and enter the correct intranet IP address.

  1. Make sure that your entered IP address of the ECS instance belongs to the logon cloud account.

  2. Currently, you can only enter the IP address of the ECS instance that is in the same region as the project. For example, if the project is in China East 1 (Hangzhou), you must enter the IP addresses of the ECS instances in the region China East 1 (Hangzhou).

  3. You must enter the intranet IP address of the ECS instance. To enter multiple IP addresses, make sure that each IP address is in a single line.

  4. You can only enter the IP address of an ECS instance, while that of a Virtual Private Cloud (VPC) instance cannot be entered.

If the issue persists, contact our after-sales technical support.

5. Different operating systems or invalid IP address. Currently, Log Service only supports ECS instances in your region

An error occurs when you add machines to a machine group in the console, indicating that the ECS instances are of different operating systems or the IP address is invalid. Currently, Log Service only supports ECS instances in your region. This error is caused because Windows ECS instances and Linux ECS instances cannot exist in the same machine group, that is, you must add Windows machines and Linux machines to different machine groups.

Complete the correct configurations and then add the machines to the machine group.

If the issue persists, contact our after-sales technical support.

6. ShardWriteQuotaExceed

This error occurs when the number of your shards is small and the write quota is exceeded. To increase the number of shards, see Split a shard in Manage a shard.

Currently, each shard has certain service capability:

  • Write: 5 MB/s, 500 times/s

  • Read: 10 MB/s, 100 times/s

Thank you! We've received your feedback.